openethereum/ethcore/snapshot/src/service.rs

918 lines
30 KiB
Rust
Raw Normal View History

// Copyright 2015-2019 Parity Technologies (UK) Ltd.
// This file is part of Parity Ethereum.
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
// Parity Ethereum is free software: you can redistribute it and/or modify
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
// Parity Ethereum is distributed in the hope that it will be useful,
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU General Public License for more details.
// You should have received a copy of the GNU General Public License
// along with Parity Ethereum. If not, see <http://www.gnu.org/licenses/>.
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
//! Snapshot network service implementation.
2017-04-19 20:31:53 +02:00
use std::collections::HashSet;
use std::io::{self, Read, ErrorKind};
use std::fs::{self, File};
use std::path::PathBuf;
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
use std::sync::Arc;
use std::sync::atomic::{AtomicBool, AtomicUsize, Ordering};
use std::cmp;
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
use blockchain::{BlockChain, BlockChainDB, BlockChainDBHandler};
use bytes::Bytes;
use common_types::{
Make ClientIoMessage generic over the Client (#10981) * Add client-traits crate Move the BlockInfo trait to new crate * New crate `machine` Contains code extracted from ethcore that defines `Machine`, `Externalities` and other execution related code. * Use new machine and client-traits crates in ethcore * Use new crates machine and client-traits instead of ethcore where appropriate * Fix tests * Don't re-export so many types from ethcore::client * Fixing more fallout from removing re-export * fix test * More fallout from not re-exporting types * Add some docs * cleanup * import the macro edition style * Tweak docs * Add missing import * remove unused ethabi_derive imports * Use latest ethabi-contract * Move many traits from ethcore/client/traits to client-traits crate Initial version of extracted Engine trait * Move snapshot related traits to the engine crate (eew) * Move a few snapshot related types to common_types Cleanup Executed as exported from machine crate * fix warning * Gradually introduce new engine crate: snapshot * ethcore typechecks with new engine crate * Sort out types outside ethcore * Add an EpochVerifier to ethash and use that in Engine.epoch_verifier() Cleanup * Document pub members * Sort out tests Sort out default impls for EpochVerifier * Add test-helpers feature and move EngineSigner impl to the right place * Sort out tests * Sort out tests and refactor verification types * Fix missing traits * More missing traits Fix Histogram * Fix tests and cleanup * cleanup * Put back needed logger import * Don't rexport common_types from ethcore/src/client Don't export ethcore::client::* * Remove files no longer used Use types from the engine crate Explicit exports from engine::engine * Get rid of itertools * Move a few more traits from ethcore to client-traits: BlockChainReset, ScheduleInfo, StateClient * Move ProvingBlockChainClient to client-traits * Don't re-export ForkChoice and Transition from ethcore * Address grumbles: sort imports, remove commented out code * Fix merge resolution error * Extract the Clique engine to own crate * Extract NullEngine and the block_reward module from ethcore * Extract InstantSeal engine to own crate * Extract remaining engines * Extract executive_state to own crate so it can be used by engine crates * Remove snapshot stuff from the engine crate * Put snapshot traits back in ethcore * cleanup * Remove stuff from ethcore * Don't use itertools * itertools in aura is legit-ish * More post-merge fixes * Re-export less types in client * cleanup * Extract spec to own crate * Put back the test-helpers from basic-authority * Fix ethcore benchmarks * Reduce the public api of ethcore/verification * WIP * Add Cargo.toml * Fix compilation outside ethcore * Audit uses of import_verified_blocks() and remove unneeded calls Cleanup * cleanup * Remove unused imports from ethcore * Cleanup * remove double semi-colons * Add missing generic param * More missing generics * Update ethcore/block-reward/Cargo.toml Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * Update ethcore/engines/basic-authority/Cargo.toml Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * Update ethcore/engines/ethash/Cargo.toml Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * Update ethcore/engines/clique/src/lib.rs Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * signers is already a ref * Add an EngineType enum to tighten up Engine.name() * Introduce Snapshotting enum to distinguish the type of snapshots a chain uses * Rename supports_warp to snapshot_mode * Missing import * Update ethcore/src/snapshot/consensus/mod.rs Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * missing import * Fix import * double semi * Fix merge problem * cleanup * Parametrise `ClientIoMessage` with `()` for the light client * Add impl Tick for () * Address review feedback * Move ClientIoMessage to common-types * remove superseeded fixme * fix merge conflict errors
2019-08-28 10:09:42 +02:00
io_message::ClientIoMessage,
Move more types out of ethcore (#10880) * WIP move errors, pod_account and state account to own crates * Sort out dependencies, fix broken code and tests Remove botched ethcore-error crate * remove template line * fix review feedback * Remove test-only AccountDBMut::new * Extract AccountDB to account-db * Move Substate to state-account – wip * Add lib.rs * cleanup * test failure * test failure 2 * third time's the charm * Add factories crate * Use new factories crate * Use factories crate * Extract trace * Fix tests * Sort out parity-util-mem and parking_lot * cleanup * WIP port over the rest of state from ethcore * Collect all impls for Machine * some notes * Rename pod-account to pod * Move PodState to pod crate * Use PodState from pod crate * Fix use clause for json tests * Sort out evmbin * Add missing code and use PodState * Move code that depends on Machine and Executive to own module * Sort out cloning errors, fix ethcore to use new state crate * Do without funky From impls * Fix ethcore tests * Fixes around the project to use new state crate * Add back the more specific impls of StateOrBlock From conversions * Move execute to freestanding function and remove it from trait Sort out the error handling in executive_state by moving the result types from state to ethcore Undo the verbose code added to work around the StateOrBlock From conversions * cleanup * Fix "error: enum variants on type aliases are experimental" * Bring back the state tests Fix whitespace * remove ethcore/state/mod.rs * cleanup * cleanup * Cleanup state-account errors * Fix more todos Add module docs * Add error.rs * Fixup Cargo.lock * Smaller ethcore API is fine * Add `to-pod-full` feature to state-account Fix evmbin * Fix a few more test failures * Fix RPC test build * Baptize the new trait * Remove resolved TODOs * Rename state-account to account-state * Do not re-export the trace crate * Don't export state_db from ethcore * Let private-tx use StateDB. :( * Remove ethcore/src/pod_state.rs * Inner type does not need to be pub/pub(crate) * optimise imports * Revert "Inner type does not need to be pub/pub(crate)" This reverts commit 2f839f8a0f72f71334da64620f57e6dd6039f06b. * Move DatabaseExtras to ethcore-blockchain * Add database_extra module to ethcore-blockchain * Remove to-pod-full feature * cosmetics * New crate: state-db * Add new crate * Move PreverifiedBlock and BlockError to types * Sort out the merge * Add missing `license` meta data keys * wip * wip client-traits * merge conflict * verification crate type checks * Move impls for CommonParams to common_types Fix misc stuff in ethcore * Fix tests * Implement VerifyingEngine for all engines except Ethash Temporarily sort out error handling Move more types to common_types * Split Engine in two and move code around * cleanup * verification: don't rexport common_types * Use error from common_types * Consolidate error types * VerifyingEngine use Errors from common_types * verification: Use error type from common_types * SnapshotError moved to common_types * Move more code from Engne to VerifyingEngine Add a VerifyingClient trait: BlockInfo + CallContract Whitespace * Add MAX_UNCLE_AGE const * Port over remaining code from ethcore/verification * Use errors from common_types * Fix the confusing "io" naming * Move more types into common_types * Add todos * Experiment with Engine trait outside ethcore * Hook up types from common_types in ethcore Don't use verification crate Don't use client-traits crate * Revert to impl Engine for Arc<Ethash> and add note to explain why Revert moving ClientIoMessage to common_types Fix build * Remove ClientIoMessage from common_types * Cleanup * More cleanup * Sort error handling changes in the rest of parity * Remove unused code * Remove WIP types * Cleanup todos not tackled here * remove cruft * Fix some whitespace and a merge error * ethcore tests * test failures * Restore Engine impls to master to make review a bit easier * cleanup * whitespace * applied review suggestions * types does not depend on rustc-hex * ethash engine moved to engine module * applied review suggestion
2019-07-18 12:27:08 +02:00
errors::{EthcoreError as Error, SnapshotError, SnapshotError::UnlinkedAncientBlockChain},
ids::BlockId,
Extract engines to own crates (#10966) * Add client-traits crate Move the BlockInfo trait to new crate * New crate `machine` Contains code extracted from ethcore that defines `Machine`, `Externalities` and other execution related code. * Use new machine and client-traits crates in ethcore * Use new crates machine and client-traits instead of ethcore where appropriate * Fix tests * Don't re-export so many types from ethcore::client * Fixing more fallout from removing re-export * fix test * More fallout from not re-exporting types * Add some docs * cleanup * import the macro edition style * Tweak docs * Add missing import * remove unused ethabi_derive imports * Use latest ethabi-contract * Move many traits from ethcore/client/traits to client-traits crate Initial version of extracted Engine trait * Move snapshot related traits to the engine crate (eew) * Move a few snapshot related types to common_types Cleanup Executed as exported from machine crate * fix warning * Gradually introduce new engine crate: snapshot * ethcore typechecks with new engine crate * Sort out types outside ethcore * Add an EpochVerifier to ethash and use that in Engine.epoch_verifier() Cleanup * Document pub members * Sort out tests Sort out default impls for EpochVerifier * Add test-helpers feature and move EngineSigner impl to the right place * Sort out tests * Sort out tests and refactor verification types * Fix missing traits * More missing traits Fix Histogram * Fix tests and cleanup * cleanup * Put back needed logger import * Don't rexport common_types from ethcore/src/client Don't export ethcore::client::* * Remove files no longer used Use types from the engine crate Explicit exports from engine::engine * Get rid of itertools * Move a few more traits from ethcore to client-traits: BlockChainReset, ScheduleInfo, StateClient * Move ProvingBlockChainClient to client-traits * Don't re-export ForkChoice and Transition from ethcore * Address grumbles: sort imports, remove commented out code * Fix merge resolution error * Extract the Clique engine to own crate * Extract NullEngine and the block_reward module from ethcore * Extract InstantSeal engine to own crate * Extract remaining engines * Extract executive_state to own crate so it can be used by engine crates * Remove snapshot stuff from the engine crate * Put snapshot traits back in ethcore * cleanup * Remove stuff from ethcore * Don't use itertools * itertools in aura is legit-ish * More post-merge fixes * Re-export less types in client * cleanup * Update ethcore/block-reward/Cargo.toml Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * Update ethcore/engines/basic-authority/Cargo.toml Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * Update ethcore/engines/ethash/Cargo.toml Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * Update ethcore/engines/clique/src/lib.rs Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * signers is already a ref * Add an EngineType enum to tighten up Engine.name() * Introduce Snapshotting enum to distinguish the type of snapshots a chain uses * Rename supports_warp to snapshot_mode * Missing import * Update ethcore/src/snapshot/consensus/mod.rs Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * remove double-semicolons
2019-08-22 18:25:49 +02:00
snapshot::{ManifestData, Progress, RestorationStatus},
Move more types out of ethcore (#10880) * WIP move errors, pod_account and state account to own crates * Sort out dependencies, fix broken code and tests Remove botched ethcore-error crate * remove template line * fix review feedback * Remove test-only AccountDBMut::new * Extract AccountDB to account-db * Move Substate to state-account – wip * Add lib.rs * cleanup * test failure * test failure 2 * third time's the charm * Add factories crate * Use new factories crate * Use factories crate * Extract trace * Fix tests * Sort out parity-util-mem and parking_lot * cleanup * WIP port over the rest of state from ethcore * Collect all impls for Machine * some notes * Rename pod-account to pod * Move PodState to pod crate * Use PodState from pod crate * Fix use clause for json tests * Sort out evmbin * Add missing code and use PodState * Move code that depends on Machine and Executive to own module * Sort out cloning errors, fix ethcore to use new state crate * Do without funky From impls * Fix ethcore tests * Fixes around the project to use new state crate * Add back the more specific impls of StateOrBlock From conversions * Move execute to freestanding function and remove it from trait Sort out the error handling in executive_state by moving the result types from state to ethcore Undo the verbose code added to work around the StateOrBlock From conversions * cleanup * Fix "error: enum variants on type aliases are experimental" * Bring back the state tests Fix whitespace * remove ethcore/state/mod.rs * cleanup * cleanup * Cleanup state-account errors * Fix more todos Add module docs * Add error.rs * Fixup Cargo.lock * Smaller ethcore API is fine * Add `to-pod-full` feature to state-account Fix evmbin * Fix a few more test failures * Fix RPC test build * Baptize the new trait * Remove resolved TODOs * Rename state-account to account-state * Do not re-export the trace crate * Don't export state_db from ethcore * Let private-tx use StateDB. :( * Remove ethcore/src/pod_state.rs * Inner type does not need to be pub/pub(crate) * optimise imports * Revert "Inner type does not need to be pub/pub(crate)" This reverts commit 2f839f8a0f72f71334da64620f57e6dd6039f06b. * Move DatabaseExtras to ethcore-blockchain * Add database_extra module to ethcore-blockchain * Remove to-pod-full feature * cosmetics * New crate: state-db * Add new crate * Move PreverifiedBlock and BlockError to types * Sort out the merge * Add missing `license` meta data keys * wip * wip client-traits * merge conflict * verification crate type checks * Move impls for CommonParams to common_types Fix misc stuff in ethcore * Fix tests * Implement VerifyingEngine for all engines except Ethash Temporarily sort out error handling Move more types to common_types * Split Engine in two and move code around * cleanup * verification: don't rexport common_types * Use error from common_types * Consolidate error types * VerifyingEngine use Errors from common_types * verification: Use error type from common_types * SnapshotError moved to common_types * Move more code from Engne to VerifyingEngine Add a VerifyingClient trait: BlockInfo + CallContract Whitespace * Add MAX_UNCLE_AGE const * Port over remaining code from ethcore/verification * Use errors from common_types * Fix the confusing "io" naming * Move more types into common_types * Add todos * Experiment with Engine trait outside ethcore * Hook up types from common_types in ethcore Don't use verification crate Don't use client-traits crate * Revert to impl Engine for Arc<Ethash> and add note to explain why Revert moving ClientIoMessage to common_types Fix build * Remove ClientIoMessage from common_types * Cleanup * More cleanup * Sort error handling changes in the rest of parity * Remove unused code * Remove WIP types * Cleanup todos not tackled here * remove cruft * Fix some whitespace and a merge error * ethcore tests * test failures * Restore Engine impls to master to make review a bit easier * cleanup * whitespace * applied review suggestions * types does not depend on rustc-hex * ethash engine moved to engine module * applied review suggestion
2019-07-18 12:27:08 +02:00
};
use client_traits::ChainInfo;
use engine::Engine;
use ethereum_types::H256;
use ethcore_io::IoChannel;
use journaldb::Algorithm;
use keccak_hash::keccak;
use kvdb::DBTransaction;
use log::{debug, error, info, trace, warn};
use parking_lot::{Mutex, RwLock, RwLockReadGuard};
use snappy;
use trie_db::TrieError;
use crate::{SnapshotClient, SnapshotWriter};
use super::{
StateRebuilder,
SnapshotService,
Rebuilder,
MAX_CHUNK_SIZE,
io::{SnapshotReader, LooseReader, LooseWriter},
chunker,
};
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
2016-09-02 19:00:20 +02:00
/// Helper for removing directories in case of error.
pub struct Guard(bool, PathBuf);
2016-09-02 19:00:20 +02:00
impl Guard {
fn new(path: PathBuf) -> Self { Guard(true, path) }
#[cfg(any(test, feature = "test-helpers"))]
pub fn benign() -> Self { Guard(false, PathBuf::default()) }
2016-09-02 19:00:20 +02:00
fn disarm(mut self) { self.0 = false }
}
impl Drop for Guard {
fn drop(&mut self) {
if self.0 {
let _ = fs::remove_dir_all(&self.1);
}
}
}
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
/// State restoration manager.
pub struct Restoration {
manifest: ManifestData,
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
state_chunks_left: HashSet<H256>,
block_chunks_left: HashSet<H256>,
state: StateRebuilder,
secondary: Box<dyn Rebuilder>,
writer: Option<LooseWriter>,
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
snappy_buffer: Bytes,
final_state_root: H256,
2016-09-02 19:00:20 +02:00
guard: Guard,
db: Arc<dyn BlockChainDB>,
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
/// Params to initialise restoration
pub struct RestorationParams<'a> {
manifest: ManifestData, // manifest to base restoration on.
pruning: Algorithm, // pruning algorithm for the database.
db: Arc<dyn BlockChainDB>, // database
writer: Option<LooseWriter>, // writer for recovered snapshot.
genesis: &'a [u8], // genesis block of the chain.
2016-09-02 19:00:20 +02:00
guard: Guard, // guard for the restoration directory.
2019-06-28 10:18:18 +02:00
engine: &'a dyn Engine,
}
#[cfg(any(test, feature = "test-helpers"))]
impl<'a> RestorationParams<'a> {
pub fn new(
manifest: ManifestData,
pruning: Algorithm,
db: Arc<dyn BlockChainDB>,
writer: Option<LooseWriter>,
genesis: &'a [u8],
guard: Guard,
engine: &'a dyn Engine,
) -> Self {
Self { manifest, pruning, db, writer, genesis, guard, engine }
}
}
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
impl Restoration {
/// Build a Restoration using the given parameters.
pub fn new(params: RestorationParams) -> Result<Self, Error> {
let manifest = params.manifest;
let state_chunks = manifest.state_hashes.iter().cloned().collect();
let block_chunks = manifest.block_hashes.iter().cloned().collect();
let raw_db = params.db;
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
let chain = BlockChain::new(Default::default(), params.genesis, raw_db.clone());
Extract engines to own crates (#10966) * Add client-traits crate Move the BlockInfo trait to new crate * New crate `machine` Contains code extracted from ethcore that defines `Machine`, `Externalities` and other execution related code. * Use new machine and client-traits crates in ethcore * Use new crates machine and client-traits instead of ethcore where appropriate * Fix tests * Don't re-export so many types from ethcore::client * Fixing more fallout from removing re-export * fix test * More fallout from not re-exporting types * Add some docs * cleanup * import the macro edition style * Tweak docs * Add missing import * remove unused ethabi_derive imports * Use latest ethabi-contract * Move many traits from ethcore/client/traits to client-traits crate Initial version of extracted Engine trait * Move snapshot related traits to the engine crate (eew) * Move a few snapshot related types to common_types Cleanup Executed as exported from machine crate * fix warning * Gradually introduce new engine crate: snapshot * ethcore typechecks with new engine crate * Sort out types outside ethcore * Add an EpochVerifier to ethash and use that in Engine.epoch_verifier() Cleanup * Document pub members * Sort out tests Sort out default impls for EpochVerifier * Add test-helpers feature and move EngineSigner impl to the right place * Sort out tests * Sort out tests and refactor verification types * Fix missing traits * More missing traits Fix Histogram * Fix tests and cleanup * cleanup * Put back needed logger import * Don't rexport common_types from ethcore/src/client Don't export ethcore::client::* * Remove files no longer used Use types from the engine crate Explicit exports from engine::engine * Get rid of itertools * Move a few more traits from ethcore to client-traits: BlockChainReset, ScheduleInfo, StateClient * Move ProvingBlockChainClient to client-traits * Don't re-export ForkChoice and Transition from ethcore * Address grumbles: sort imports, remove commented out code * Fix merge resolution error * Extract the Clique engine to own crate * Extract NullEngine and the block_reward module from ethcore * Extract InstantSeal engine to own crate * Extract remaining engines * Extract executive_state to own crate so it can be used by engine crates * Remove snapshot stuff from the engine crate * Put snapshot traits back in ethcore * cleanup * Remove stuff from ethcore * Don't use itertools * itertools in aura is legit-ish * More post-merge fixes * Re-export less types in client * cleanup * Update ethcore/block-reward/Cargo.toml Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * Update ethcore/engines/basic-authority/Cargo.toml Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * Update ethcore/engines/ethash/Cargo.toml Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * Update ethcore/engines/clique/src/lib.rs Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * signers is already a ref * Add an EngineType enum to tighten up Engine.name() * Introduce Snapshotting enum to distinguish the type of snapshots a chain uses * Rename supports_warp to snapshot_mode * Missing import * Update ethcore/src/snapshot/consensus/mod.rs Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * remove double-semicolons
2019-08-22 18:25:49 +02:00
let chunker = chunker(params.engine.snapshot_mode())
.ok_or_else(|| Error::Snapshot(SnapshotError::SnapshotsUnsupported))?;
2017-04-19 20:31:53 +02:00
Extract engines to own crates (#10966) * Add client-traits crate Move the BlockInfo trait to new crate * New crate `machine` Contains code extracted from ethcore that defines `Machine`, `Externalities` and other execution related code. * Use new machine and client-traits crates in ethcore * Use new crates machine and client-traits instead of ethcore where appropriate * Fix tests * Don't re-export so many types from ethcore::client * Fixing more fallout from removing re-export * fix test * More fallout from not re-exporting types * Add some docs * cleanup * import the macro edition style * Tweak docs * Add missing import * remove unused ethabi_derive imports * Use latest ethabi-contract * Move many traits from ethcore/client/traits to client-traits crate Initial version of extracted Engine trait * Move snapshot related traits to the engine crate (eew) * Move a few snapshot related types to common_types Cleanup Executed as exported from machine crate * fix warning * Gradually introduce new engine crate: snapshot * ethcore typechecks with new engine crate * Sort out types outside ethcore * Add an EpochVerifier to ethash and use that in Engine.epoch_verifier() Cleanup * Document pub members * Sort out tests Sort out default impls for EpochVerifier * Add test-helpers feature and move EngineSigner impl to the right place * Sort out tests * Sort out tests and refactor verification types * Fix missing traits * More missing traits Fix Histogram * Fix tests and cleanup * cleanup * Put back needed logger import * Don't rexport common_types from ethcore/src/client Don't export ethcore::client::* * Remove files no longer used Use types from the engine crate Explicit exports from engine::engine * Get rid of itertools * Move a few more traits from ethcore to client-traits: BlockChainReset, ScheduleInfo, StateClient * Move ProvingBlockChainClient to client-traits * Don't re-export ForkChoice and Transition from ethcore * Address grumbles: sort imports, remove commented out code * Fix merge resolution error * Extract the Clique engine to own crate * Extract NullEngine and the block_reward module from ethcore * Extract InstantSeal engine to own crate * Extract remaining engines * Extract executive_state to own crate so it can be used by engine crates * Remove snapshot stuff from the engine crate * Put snapshot traits back in ethcore * cleanup * Remove stuff from ethcore * Don't use itertools * itertools in aura is legit-ish * More post-merge fixes * Re-export less types in client * cleanup * Update ethcore/block-reward/Cargo.toml Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * Update ethcore/engines/basic-authority/Cargo.toml Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * Update ethcore/engines/ethash/Cargo.toml Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * Update ethcore/engines/clique/src/lib.rs Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * signers is already a ref * Add an EngineType enum to tighten up Engine.name() * Introduce Snapshotting enum to distinguish the type of snapshots a chain uses * Rename supports_warp to snapshot_mode * Missing import * Update ethcore/src/snapshot/consensus/mod.rs Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * remove double-semicolons
2019-08-22 18:25:49 +02:00
let secondary = chunker.rebuilder(chain, raw_db.clone(), &manifest)?;
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
let final_state_root = manifest.state_root.clone();
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
Ok(Restoration {
manifest,
state_chunks_left: state_chunks,
block_chunks_left: block_chunks,
state: StateRebuilder::new(raw_db.key_value().clone(), params.pruning),
secondary,
writer: params.writer,
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
snappy_buffer: Vec::new(),
final_state_root,
2016-09-02 19:00:20 +02:00
guard: params.guard,
db: raw_db,
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
})
}
/// Feeds a chunk of state data to the Restoration. Aborts early if `flag` becomes false.
pub fn feed_state(&mut self, hash: H256, chunk: &[u8], flag: &AtomicBool) -> Result<(), Error> {
if self.state_chunks_left.contains(&hash) {
let expected_len = snappy::decompressed_len(chunk)?;
if expected_len > MAX_CHUNK_SIZE {
trace!(target: "snapshot", "Discarding large chunk: {} vs {}", expected_len, MAX_CHUNK_SIZE);
return Err(SnapshotError::ChunkTooLarge.into());
}
let len = snappy::decompress_into(chunk, &mut self.snappy_buffer)?;
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
self.state.feed(&self.snappy_buffer[..len], flag)?;
if let Some(ref mut writer) = self.writer.as_mut() {
writer.write_state_chunk(hash, chunk)?;
Snapshot restoration overhaul (#11219) * Comments and todos Use `snapshot_sync` as logging target * fix compilation * More todos, more logs * Fix picking snapshot peer: prefer the one with the highest block number More docs, comments, todos * Adjust WAIT_PEERS_TIMEOUT to be a multiple of MAINTAIN_SYNC_TIMER to try to fix snapshot startup problems Docs, todos, comments * Tabs * Formatting * Don't build new rlp::EMPTY_LIST_RLP instances * Dial down debug logging * Don't warn about missing hashes in the manifest: it's normal Log client version on peer connect * Cleanup * Do not skip snapshots further away than 30k block from the highest block seen Currently we look for peers that seed snapshots that are close to the highest block seen on the network (where "close" means withing 30k blocks). When a node starts up we wait for some time (5sec, increased here to 10sec) to let peers connect and if we have found a suitable peer to sync a snapshot from at the end of that delay, we start the download; if none is found and --warp-barrier is used we stall, otherwise we start a slow-sync. When looking for a suitable snapshot, we use the highest block seen on the network to check if a peer has a snapshot that is within 30k blocks of that highest block number. This means that in a situation where all available snapshots are older than that, we will often fail to start a snapshot at all. What's worse is that the longer we delay starting a snapshot sync (to let more peers connect, in the hope of finding a good snapshot), the more likely we are to have seen a high block and thus the more likely we become to accept a snapshot. This commit removes this comparison with the highest blocknumber criteria entirely and picks the best snapshot we find in 10sec. * lockfile * Add a `ChunkType::Dupe` variant so that we do not disconnect a peer if they happen to send us a duplicate chunk (just ignore the chunk and keep going) Resolve some documentation todos, add more * tweak log message * Don't warp sync twice Check if our own block is beyond the given warp barrier (can happen after we've completed a warp sync but are not quite yet synced up to the tip) and if so, don't sync. More docs, resolve todos. Dial down some `sync` debug level logging to trace * Avoid iterating over all snapshot block/state hashes to find the next work item Use a HashSet instead of a Vec and remove items from the set as chunks are processed. Calculate and store the total number of chunks in the `Snapshot` struct instead of counting pending chunks each time. * Address review grumbles * Log correct number of bytes written to disk * Revert ChunkType::Dup change * whitespace grumble * Cleanup debugging code * Fix docs * Fix import and a typo * Fix test impl * Use `indexmap::IndexSet` to ensure chunk hashes are accessed in order * Revert increased SNAPSHOT_MANIFEST_TIMEOUT: 5sec should be enough
2019-10-31 16:07:21 +01:00
trace!(target: "snapshot", "Wrote {}/{} bytes of state to db/disk. Current state root: {:?}", len, chunk.len(), self.state.state_root());
}
self.state_chunks_left.remove(&hash);
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
Ok(())
}
/// Feeds a chunk of block data to the `Restoration`. Aborts early if `flag` becomes false.
pub fn feed_blocks(&mut self, hash: H256, chunk: &[u8], engine: &dyn Engine, flag: &AtomicBool) -> Result<(), Error> {
if self.block_chunks_left.contains(&hash) {
let expected_len = snappy::decompressed_len(chunk)?;
if expected_len > MAX_CHUNK_SIZE {
trace!(target: "snapshot", "Discarding large chunk: {} vs {}", expected_len, MAX_CHUNK_SIZE);
return Err(SnapshotError::ChunkTooLarge.into());
}
let len = snappy::decompress_into(chunk, &mut self.snappy_buffer)?;
2017-04-19 20:31:53 +02:00
self.secondary.feed(&self.snappy_buffer[..len], engine, flag)?;
if let Some(ref mut writer) = self.writer.as_mut() {
writer.write_block_chunk(hash, chunk)?;
}
self.block_chunks_left.remove(&hash);
}
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
Ok(())
}
// finish up restoration.
fn finalize(mut self) -> Result<(), Error> {
if !self.is_done() { return Ok(()) }
// verify final state root.
let root = self.state.state_root();
if root != self.final_state_root {
warn!("Final restored state has wrong state root: expected {:?}, got {:?}", self.final_state_root, root);
return Err(TrieError::InvalidStateRoot(root).into());
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
// check for missing code.
self.state.finalize(self.manifest.block_number, self.manifest.block_hash)?;
// connect out-of-order chunks and verify chain integrity.
self.secondary.finalize()?;
if let Some(writer) = self.writer {
writer.finish(self.manifest)?;
}
2016-09-02 19:00:20 +02:00
self.guard.disarm();
trace!(target: "snapshot", "Restoration finalised correctly");
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
Ok(())
}
/// Check if we're done restoring: no more block chunks and no more state chunks to process.
pub fn is_done(&self) -> bool {
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
self.block_chunks_left.is_empty() && self.state_chunks_left.is_empty()
}
}
/// Type alias for client io channel.
pub type Channel<C> = IoChannel<ClientIoMessage<C>>;
/// Snapshot service parameters.
pub struct ServiceParams<C: 'static> {
/// The consensus engine this is built on.
2019-06-28 10:18:18 +02:00
pub engine: Arc<dyn Engine>,
/// The chain's genesis block.
pub genesis_block: Bytes,
/// State pruning algorithm.
pub pruning: Algorithm,
/// Handler for opening a restoration DB.
pub restoration_db_handler: Box<dyn BlockChainDBHandler>,
/// Async IO channel for sending messages.
pub channel: Channel<C>,
/// The directory to put snapshots in.
/// Usually "<chain hash>/snapshot"
pub snapshot_root: PathBuf,
/// A handle for database restoration.
pub client: Arc<C>,
}
/// `SnapshotService` implementation.
/// This controls taking snapshots and restoring from them.
pub struct Service<C: Send + Sync + 'static> {
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
restoration: Mutex<Option<Restoration>>,
restoration_db_handler: Box<dyn BlockChainDBHandler>,
snapshot_root: PathBuf,
io_channel: Mutex<Channel<C>>,
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
pruning: Algorithm,
status: Mutex<RestorationStatus>,
reader: RwLock<Option<LooseReader>>,
2019-06-28 10:18:18 +02:00
engine: Arc<dyn Engine>,
genesis_block: Bytes,
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
state_chunks: AtomicUsize,
block_chunks: AtomicUsize,
client: Arc<C>,
progress: RwLock<Progress>,
taking_snapshot: AtomicBool,
restoring_snapshot: AtomicBool,
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
impl<C> Service<C> where C: SnapshotClient + ChainInfo {
/// Create a new snapshot service from the given parameters.
pub fn new(params: ServiceParams<C>) -> Result<Self, Error> {
let mut service = Service {
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
restoration: Mutex::new(None),
restoration_db_handler: params.restoration_db_handler,
snapshot_root: params.snapshot_root,
2016-10-30 09:56:34 +01:00
io_channel: Mutex::new(params.channel),
pruning: params.pruning,
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
status: Mutex::new(RestorationStatus::Inactive),
reader: RwLock::new(None),
engine: params.engine,
genesis_block: params.genesis_block,
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
state_chunks: AtomicUsize::new(0),
block_chunks: AtomicUsize::new(0),
client: params.client,
progress: RwLock::new(Progress::new()),
taking_snapshot: AtomicBool::new(false),
restoring_snapshot: AtomicBool::new(false),
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
};
// create the root snapshot dir if it doesn't exist.
if let Err(e) = fs::create_dir_all(&service.snapshot_root) {
if e.kind() != ErrorKind::AlreadyExists {
return Err(e.into())
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
}
// delete the temporary restoration DB dir if it does exist.
if let Err(e) = fs::remove_dir_all(service.restoration_db()) {
if e.kind() != ErrorKind::NotFound {
return Err(e.into())
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
}
2016-09-05 14:25:56 +02:00
// delete the temporary snapshot dir if it does exist.
if let Err(e) = fs::remove_dir_all(service.temp_snapshot_dir()) {
if e.kind() != ErrorKind::NotFound {
2016-09-05 14:25:56 +02:00
return Err(e.into())
}
}
let reader = LooseReader::new(service.snapshot_dir()).ok();
*service.reader.get_mut() = reader;
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
Ok(service)
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
// get the current snapshot dir.
fn snapshot_dir(&self) -> PathBuf {
let mut dir = self.snapshot_root.clone();
dir.push("current");
dir
}
2016-09-02 16:15:25 +02:00
// get the temporary snapshot dir.
fn temp_snapshot_dir(&self) -> PathBuf {
let mut dir = self.snapshot_root.clone();
2016-09-02 16:15:25 +02:00
dir.push("in_progress");
dir
}
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
// get the restoration directory.
fn restoration_dir(&self) -> PathBuf {
let mut dir = self.snapshot_root.clone();
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
dir.push("restoration");
dir
}
// restoration db path.
fn restoration_db(&self) -> PathBuf {
let mut dir = self.restoration_dir();
dir.push("db");
dir
}
// temporary snapshot recovery path.
fn temp_recovery_dir(&self) -> PathBuf {
let mut dir = self.restoration_dir();
dir.push("temp");
dir
}
// previous snapshot chunks path.
fn prev_chunks_dir(&self) -> PathBuf {
let mut dir = self.snapshot_root.clone();
dir.push("prev_chunks");
dir
}
// Migrate the blocks in the current DB into the new chain
fn migrate_blocks(&self) -> Result<usize, Error> {
// Count the number of migrated blocks
let mut count = 0;
let rest_db = self.restoration_db();
let cur_chain_info = self.client.chain_info();
let next_db = self.restoration_db_handler.open(&rest_db)?;
let next_chain = BlockChain::new(Default::default(), &[], next_db.clone());
let next_chain_info = next_chain.chain_info();
// The old database looks like this:
// [genesis, best_ancient_block] ... [first_block, best_block]
// If we are fully synced neither `best_ancient_block` nor `first_block` is set, and we can
// assume that the whole range from [genesis, best_block] is imported.
// The new database only contains the tip of the chain ([new_first_block, new_best_block]),
// so the useful set of blocks is defined as:
// [0 ... min(new.first_block, best_ancient_block or best_block)]
//
// If, for whatever reason, the old db does not have ancient blocks (i.e.
// `best_ancient_block` is `None` AND a non-zero `first_block`), such that the old db looks
// like [old_first_block..old_best_block] (which may or may not partially overlap with
// [new_first_block..new_best_block]) we do the conservative thing and do not migrate the
// old blocks.
let find_range = || -> Option<(H256, H256)> {
// In theory, if the current best_block is > new first_block (i.e. ranges overlap)
// we could salvage them but what if there's been a re-org at the boundary and the two
// chains do not match anymore? We'd have to check the existing blocks carefully.
if cur_chain_info.ancient_block_number.is_none() && cur_chain_info.first_block_number.unwrap_or(0) > 0 {
info!(target: "blockchain", "blocks in the current DB do not stretch back to genesis; can't salvage them into the new DB. In current DB, first block: #{:?}/{:?}, best block: #{:?}/{:?}",
cur_chain_info.first_block_number, cur_chain_info.first_block_hash,
cur_chain_info.best_block_number, cur_chain_info.best_block_hash);
return None;
}
let next_available_from = next_chain_info.first_block_number?;
let cur_available_to = cur_chain_info.ancient_block_number.unwrap_or(cur_chain_info.best_block_number);
let highest_block_num = cmp::min(next_available_from.saturating_sub(1), cur_available_to);
if highest_block_num == 0 {
return None;
}
trace!(target: "snapshot", "Trying to import ancient blocks until {}. First block in new chain=#{}, first block in old chain=#{:?}, best block in old chain=#{}",
highest_block_num, next_available_from, cur_chain_info.first_block_number, cur_chain_info.best_block_number);
// Here we start from the highest block number and go backward to 0,
// thus starting at `highest_block_num` and targeting `0`.
let target_hash = self.client.block_hash(BlockId::Number(0))?;
let start_hash = self.client.block_hash(BlockId::Number(highest_block_num))?;
Some((start_hash, target_hash))
};
let (start_hash, target_hash) = match find_range() {
Some(x) => x,
None => return Ok(0),
};
info!(target: "snapshot", "Migrating blocks from old db to new. Start: #{}/{:?}, Target: #{}/{:?}",
self.client.block_number(BlockId::Hash(start_hash)).unwrap_or_default(), start_hash,
self.client.block_number(BlockId::Hash(target_hash)).unwrap_or_default(), target_hash,
);
let mut batch = DBTransaction::new();
let mut parent_hash = start_hash;
while parent_hash != target_hash {
// Early return if restoration is aborted
if !self.restoring_snapshot.load(Ordering::SeqCst) {
return Ok(count);
}
let block = self.client.block(BlockId::Hash(parent_hash)).ok_or_else(|| {
error!(target: "snapshot", "migrate_blocks: did not find block from parent_hash={:#x} (start_hash={:#x})", parent_hash, start_hash);
UnlinkedAncientBlockChain(parent_hash)
})?;
parent_hash = block.parent_hash();
let block_number = block.number();
let block_receipts = self.client.block_receipts(&block.hash());
let parent_total_difficulty = self.client.block_total_difficulty(BlockId::Hash(parent_hash));
match (block_receipts, parent_total_difficulty) {
(Some(block_receipts), Some(parent_total_difficulty)) => {
let block_receipts = block_receipts.receipts;
next_chain.insert_unordered_block(&mut batch, block, block_receipts, Some(parent_total_difficulty), false, true);
count += 1;
},
_ => {
// We couldn't reach the targeted hash
error!(target: "snapshot", "migrate_blocks: failed to find receipts and parent total difficulty; cannot reach the target_hash ({:#x}). Block #{}, parent_hash={:#x}, parent_total_difficulty={:?}, start_hash={:#x}, ancient_block_number={:?}, best_block_number={:?}",
target_hash, block_number, parent_hash, parent_total_difficulty,
start_hash, cur_chain_info.ancient_block_number, cur_chain_info.best_block_number,
);
return Err(UnlinkedAncientBlockChain(parent_hash).into());
},
}
// Writing changes to DB and logging every now and then
if block_number % 1_000 == 0 {
next_db.key_value().write_buffered(batch);
next_chain.commit();
next_db.key_value().flush().expect("DB flush failed.");
batch = DBTransaction::new();
if block_number % 10_000 == 0 {
info!(target: "snapshot", "Block restoration at #{}", block_number);
}
}
}
// Final commit to the DB
next_db.key_value().write_buffered(batch);
next_chain.commit();
next_db.key_value().flush().expect("DB flush failed.");
// Update best ancient block in the Next Chain
next_chain.update_best_ancient_block(&start_hash);
Ok(count)
}
/// Get a reference to the snapshot reader.
pub fn reader(&self) -> RwLockReadGuard<Option<LooseReader>> {
self.reader.read()
}
2016-09-06 17:44:11 +02:00
/// Tick the snapshot service. This will log any active snapshot
/// being taken.
pub fn tick(&self) {
if self.progress.read().done() || !self.taking_snapshot.load(Ordering::SeqCst) { return }
2016-09-06 17:44:11 +02:00
let p = &self.progress.read();
info!("Snapshot: {} accounts, {} blocks, {} bytes", p.accounts(), p.blocks(), p.bytes());
let rate = p.rate();
debug!(target: "snapshot", "Current progress rate: {:.0} acc/s, {:.0} bytes/s (compressed)", rate.0, rate.1);
2016-09-06 17:44:11 +02:00
}
2016-09-02 16:15:25 +02:00
/// Take a snapshot at the block with the given number.
/// Calling this while a restoration is in progress or vice versa
2016-09-02 16:15:25 +02:00
/// will lead to a race condition where the first one to finish will
/// have their produced snapshot overwritten.
pub fn take_snapshot(&self, client: &C, num: u64) -> Result<(), Error> {
if self.taking_snapshot.compare_and_swap(false, true, Ordering::SeqCst) {
info!("Skipping snapshot at #{} as another one is currently in-progress.", num);
return Ok(());
}
2016-09-02 16:15:25 +02:00
info!("Taking snapshot at #{}", num);
{
scopeguard::defer! {{
self.taking_snapshot.store(false, Ordering::SeqCst);
}}
let start_time = std::time::Instant::now();
*self.progress.write() = Progress::new();
let temp_dir = self.temp_snapshot_dir();
let snapshot_dir = self.snapshot_dir();
2016-09-02 16:15:25 +02:00
let _ = fs::remove_dir_all(&temp_dir); // expected to fail
2016-09-02 16:15:25 +02:00
let writer = LooseWriter::new(temp_dir.clone())?;
2016-09-02 19:00:20 +02:00
let guard = Guard::new(temp_dir.clone());
let _ = client.take_snapshot(writer, BlockId::Number(num), &self.progress)?;
info!("Finished taking snapshot at #{}, in {:.0?}", num, start_time.elapsed());
2016-09-02 16:15:25 +02:00
// destroy the old snapshot reader.
let mut reader = self.reader.write();
*reader = None;
2016-09-06 17:44:11 +02:00
if snapshot_dir.exists() {
trace!(target: "snapshot", "Removing previous snapshot at {:?}", &snapshot_dir);
fs::remove_dir_all(&snapshot_dir)?;
}
2016-09-02 16:15:25 +02:00
fs::rename(temp_dir, &snapshot_dir)?;
trace!(target: "snapshot", "Moved new snapshot into place at {:?}", &snapshot_dir);
*reader = Some(LooseReader::new(snapshot_dir)?);
2016-09-02 16:15:25 +02:00
guard.disarm();
Ok(())
}
2016-09-02 16:15:25 +02:00
}
/// Initialize the restoration synchronously.
/// The recover flag indicates whether to recover the restored snapshot.
pub fn init_restore(&self, manifest: ManifestData, recover: bool) -> Result<(), Error> {
let mut res = self.restoration.lock();
let rest_dir = self.restoration_dir();
let rest_db = self.restoration_db();
let recovery_temp = self.temp_recovery_dir();
let prev_chunks = self.prev_chunks_dir();
// delete and restore the restoration dir.
if let Err(e) = fs::remove_dir_all(&prev_chunks) {
match e.kind() {
ErrorKind::NotFound => {},
_ => return Err(e.into()),
}
}
// Move the previous recovery temp directory
// to `prev_chunks` to be able to restart restoring
// with previously downloaded blocks
// This step is optional, so don't fail on error
fs::rename(&recovery_temp, &prev_chunks).ok();
self.state_chunks.store(0, Ordering::SeqCst);
self.block_chunks.store(0, Ordering::SeqCst);
// tear down existing restoration.
*res = None;
// delete and restore the restoration dir.
if let Err(e) = fs::remove_dir_all(&rest_dir) {
match e.kind() {
ErrorKind::NotFound => {},
_ => return Err(e.into()),
}
}
*self.status.lock() = RestorationStatus::Initializing {
chunks_done: 0,
state_chunks: manifest.state_hashes.len() as u32,
block_chunks: manifest.block_hashes.len() as u32,
};
fs::create_dir_all(&rest_dir)?;
// make new restoration.
let writer = match recover {
true => Some(LooseWriter::new(recovery_temp)?),
false => None
};
let params = RestorationParams {
manifest: manifest.clone(),
pruning: self.pruning,
db: self.restoration_db_handler.open(&rest_db)?,
writer,
genesis: &self.genesis_block,
guard: Guard::new(rest_db),
2017-04-19 20:31:53 +02:00
engine: &*self.engine,
};
let state_chunks = manifest.state_hashes.len();
let block_chunks = manifest.block_hashes.len();
*res = Some(Restoration::new(params)?);
self.restoring_snapshot.store(true, Ordering::SeqCst);
// Import previous chunks, continue if it fails
self.import_prev_chunks(&mut res, manifest).ok();
// It could be that the restoration failed or completed in the meanwhile
let mut restoration_status = self.status.lock();
if let RestorationStatus::Initializing { .. } = *restoration_status {
*restoration_status = RestorationStatus::Ongoing {
state_chunks: state_chunks as u32,
block_chunks: block_chunks as u32,
state_chunks_done: self.state_chunks.load(Ordering::SeqCst) as u32,
block_chunks_done: self.block_chunks.load(Ordering::SeqCst) as u32,
};
}
Ok(())
}
/// Import the previous chunks into the current restoration
fn import_prev_chunks(&self, restoration: &mut Option<Restoration>, manifest: ManifestData) -> Result<(), Error> {
let prev_chunks = self.prev_chunks_dir();
// Restore previous snapshot chunks
let files = fs::read_dir(prev_chunks.as_path())?;
let mut num_temp_chunks = 0;
for prev_chunk_file in files {
// Don't go over all the files if the restoration has been aborted
if !self.restoring_snapshot.load(Ordering::SeqCst) {
trace!(target:"snapshot", "Aborting importing previous chunks");
return Ok(());
}
// Import the chunk, don't fail and continue if one fails
match self.import_prev_chunk(restoration, &manifest, prev_chunk_file) {
Ok(true) => num_temp_chunks += 1,
Err(e) => trace!(target: "snapshot", "Error importing chunk: {:?}", e),
_ => (),
}
}
trace!(target:"snapshot", "Imported {} previous chunks", num_temp_chunks);
// Remove the prev temp directory
fs::remove_dir_all(&prev_chunks)?;
Ok(())
}
/// Import a previous chunk at the given path. Returns whether the chunk was imported or not
fn import_prev_chunk(
&self,
restoration: &mut Option<Restoration>,
manifest: &ManifestData,
file: io::Result<fs::DirEntry>
) -> Result<bool, Error> {
let file = file?;
let path = file.path();
let mut file = File::open(path.clone())?;
let filesize = file.metadata()?.len();
let mut buffer = Vec::with_capacity(filesize as usize + 1); // +1 for EOF
file.read_to_end(&mut buffer)?;
let hash = keccak(&buffer);
let is_state = if manifest.block_hashes.contains(&hash) {
false
} else if manifest.state_hashes.contains(&hash) {
true
} else {
return Ok(false);
};
self.feed_chunk_with_restoration(restoration, hash, &buffer, is_state)?;
trace!(target: "snapshot", "Fed chunk {:?}", hash);
Ok(true)
}
// Finalize the restoration. This accepts an already-locked restoration as an argument -- so
// acquiring it again _will_ lead to deadlock.
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
fn finalize_restoration(&self, rest: &mut Option<Restoration>) -> Result<(), Error> {
trace!(target: "snapshot", "Finalizing restoration");
*self.status.lock() = RestorationStatus::Finalizing;
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
let recover = rest.as_ref().map_or(false, |rest| rest.writer.is_some());
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
// destroy the restoration before replacing databases and snapshot.
rest.take()
.map(|r| r.finalize())
.unwrap_or(Ok(()))?;
let migrated_blocks = self.migrate_blocks()?;
info!(target: "snapshot", "Migrated {} ancient blocks from the old DB", migrated_blocks);
// replace the Client's database with the new one (restart the Client).
self.client.restore_db(&*self.restoration_db().to_string_lossy())?;
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
if recover {
let mut reader = self.reader.write();
*reader = None; // destroy the old reader if it existed.
let snapshot_dir = self.snapshot_dir();
if snapshot_dir.exists() {
trace!(target: "snapshot", "Removing old snapshot dir at {}", snapshot_dir.to_string_lossy());
fs::remove_dir_all(&snapshot_dir)?;
}
trace!(target: "snapshot", "Copying restored snapshot files over");
fs::rename(self.temp_recovery_dir(), &snapshot_dir)?;
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
*reader = Some(LooseReader::new(snapshot_dir)?);
}
let _ = fs::remove_dir_all(self.restoration_dir());
*self.status.lock() = RestorationStatus::Inactive;
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
Ok(())
}
/// Feed a chunk of either kind (block or state). no-op if no restoration or status is wrong.
fn feed_chunk(&self, hash: H256, chunk: &[u8], is_state: bool) {
// TODO: be able to process block chunks and state chunks at same time?
let r = {
let mut restoration = self.restoration.lock();
self.feed_chunk_with_restoration(&mut restoration, hash, chunk, is_state)
};
match r {
Ok(()) |
Err(Error::Snapshot(SnapshotError::RestorationAborted)) => (),
Err(e) => {
// TODO: after this we're sometimes deadlocked
warn!("Encountered error during snapshot restoration: {}", e);
self.abort_restore();
if let Some(mut status) = self.status.try_lock_for(std::time::Duration::from_millis(10)) {
*status = RestorationStatus::Failed;
}
let _ = fs::remove_dir_all(self.restoration_dir());
}
}
}
/// Feed a chunk with the Restoration
fn feed_chunk_with_restoration(&self, restoration: &mut Option<Restoration>, hash: H256, chunk: &[u8], is_state: bool) -> Result<(), Error> {
let (result, db) = {
match self.status() {
RestorationStatus::Inactive | RestorationStatus::Failed | RestorationStatus::Finalizing => {
trace!(target: "snapshot", "Tried to restore chunk {:x} while inactive, failed or finalizing", hash);
return Ok(());
},
RestorationStatus::Ongoing { .. } | RestorationStatus::Initializing { .. } => {
let (res, db) = {
let rest = match *restoration {
Some(ref mut r) => r,
None => return Ok(()),
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
};
(match is_state {
true => rest.feed_state(hash, chunk, &self.restoring_snapshot),
false => rest.feed_blocks(hash, chunk, &*self.engine, &self.restoring_snapshot),
}.map(|_| rest.is_done()), rest.db.clone())
};
let res = match res {
Ok(is_done) => {
match is_state {
true => self.state_chunks.fetch_add(1, Ordering::SeqCst),
false => self.block_chunks.fetch_add(1, Ordering::SeqCst),
};
match is_done {
true => {
db.key_value().flush()?;
drop(db);
return self.finalize_restoration(&mut *restoration);
},
false => Ok(())
}
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
Snapshot restoration overhaul (#11219) * Comments and todos Use `snapshot_sync` as logging target * fix compilation * More todos, more logs * Fix picking snapshot peer: prefer the one with the highest block number More docs, comments, todos * Adjust WAIT_PEERS_TIMEOUT to be a multiple of MAINTAIN_SYNC_TIMER to try to fix snapshot startup problems Docs, todos, comments * Tabs * Formatting * Don't build new rlp::EMPTY_LIST_RLP instances * Dial down debug logging * Don't warn about missing hashes in the manifest: it's normal Log client version on peer connect * Cleanup * Do not skip snapshots further away than 30k block from the highest block seen Currently we look for peers that seed snapshots that are close to the highest block seen on the network (where "close" means withing 30k blocks). When a node starts up we wait for some time (5sec, increased here to 10sec) to let peers connect and if we have found a suitable peer to sync a snapshot from at the end of that delay, we start the download; if none is found and --warp-barrier is used we stall, otherwise we start a slow-sync. When looking for a suitable snapshot, we use the highest block seen on the network to check if a peer has a snapshot that is within 30k blocks of that highest block number. This means that in a situation where all available snapshots are older than that, we will often fail to start a snapshot at all. What's worse is that the longer we delay starting a snapshot sync (to let more peers connect, in the hope of finding a good snapshot), the more likely we are to have seen a high block and thus the more likely we become to accept a snapshot. This commit removes this comparison with the highest blocknumber criteria entirely and picks the best snapshot we find in 10sec. * lockfile * Add a `ChunkType::Dupe` variant so that we do not disconnect a peer if they happen to send us a duplicate chunk (just ignore the chunk and keep going) Resolve some documentation todos, add more * tweak log message * Don't warp sync twice Check if our own block is beyond the given warp barrier (can happen after we've completed a warp sync but are not quite yet synced up to the tip) and if so, don't sync. More docs, resolve todos. Dial down some `sync` debug level logging to trace * Avoid iterating over all snapshot block/state hashes to find the next work item Use a HashSet instead of a Vec and remove items from the set as chunks are processed. Calculate and store the total number of chunks in the `Snapshot` struct instead of counting pending chunks each time. * Address review grumbles * Log correct number of bytes written to disk * Revert ChunkType::Dup change * whitespace grumble * Cleanup debugging code * Fix docs * Fix import and a typo * Fix test impl * Use `indexmap::IndexSet` to ensure chunk hashes are accessed in order * Revert increased SNAPSHOT_MANIFEST_TIMEOUT: 5sec should be enough
2019-10-31 16:07:21 +01:00
Err(e) => Err(e)
};
(res, db)
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
}
};
result?;
db.key_value().flush()?;
Ok(())
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
/// Feed a state chunk to be processed synchronously.
pub fn feed_state_chunk(&self, hash: H256, chunk: &[u8]) {
self.feed_chunk(hash, chunk, true);
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
/// Feed a block chunk to be processed synchronously.
pub fn feed_block_chunk(&self, hash: H256, chunk: &[u8]) {
self.feed_chunk(hash, chunk, false);
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
}
impl<C: Send + Sync> SnapshotService for Service<C> {
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
fn manifest(&self) -> Option<ManifestData> {
self.reader.read().as_ref().map(|r| r.manifest().clone())
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
fn supported_versions(&self) -> Option<(u64, u64)> {
Extract engines to own crates (#10966) * Add client-traits crate Move the BlockInfo trait to new crate * New crate `machine` Contains code extracted from ethcore that defines `Machine`, `Externalities` and other execution related code. * Use new machine and client-traits crates in ethcore * Use new crates machine and client-traits instead of ethcore where appropriate * Fix tests * Don't re-export so many types from ethcore::client * Fixing more fallout from removing re-export * fix test * More fallout from not re-exporting types * Add some docs * cleanup * import the macro edition style * Tweak docs * Add missing import * remove unused ethabi_derive imports * Use latest ethabi-contract * Move many traits from ethcore/client/traits to client-traits crate Initial version of extracted Engine trait * Move snapshot related traits to the engine crate (eew) * Move a few snapshot related types to common_types Cleanup Executed as exported from machine crate * fix warning * Gradually introduce new engine crate: snapshot * ethcore typechecks with new engine crate * Sort out types outside ethcore * Add an EpochVerifier to ethash and use that in Engine.epoch_verifier() Cleanup * Document pub members * Sort out tests Sort out default impls for EpochVerifier * Add test-helpers feature and move EngineSigner impl to the right place * Sort out tests * Sort out tests and refactor verification types * Fix missing traits * More missing traits Fix Histogram * Fix tests and cleanup * cleanup * Put back needed logger import * Don't rexport common_types from ethcore/src/client Don't export ethcore::client::* * Remove files no longer used Use types from the engine crate Explicit exports from engine::engine * Get rid of itertools * Move a few more traits from ethcore to client-traits: BlockChainReset, ScheduleInfo, StateClient * Move ProvingBlockChainClient to client-traits * Don't re-export ForkChoice and Transition from ethcore * Address grumbles: sort imports, remove commented out code * Fix merge resolution error * Extract the Clique engine to own crate * Extract NullEngine and the block_reward module from ethcore * Extract InstantSeal engine to own crate * Extract remaining engines * Extract executive_state to own crate so it can be used by engine crates * Remove snapshot stuff from the engine crate * Put snapshot traits back in ethcore * cleanup * Remove stuff from ethcore * Don't use itertools * itertools in aura is legit-ish * More post-merge fixes * Re-export less types in client * cleanup * Update ethcore/block-reward/Cargo.toml Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * Update ethcore/engines/basic-authority/Cargo.toml Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * Update ethcore/engines/ethash/Cargo.toml Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * Update ethcore/engines/clique/src/lib.rs Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * signers is already a ref * Add an EngineType enum to tighten up Engine.name() * Introduce Snapshotting enum to distinguish the type of snapshots a chain uses * Rename supports_warp to snapshot_mode * Missing import * Update ethcore/src/snapshot/consensus/mod.rs Co-Authored-By: Tomasz Drwięga <tomusdrw@users.noreply.github.com> * remove double-semicolons
2019-08-22 18:25:49 +02:00
chunker(self.engine.snapshot_mode())
.map(|c| (c.min_supported_version(), c.current_version()))
}
fn completed_chunks(&self) -> Option<Vec<H256>> {
let restoration = self.restoration.lock();
match *restoration {
Some(ref restoration) => {
let completed_chunks = restoration.manifest.block_hashes
.iter()
.filter(|h| !restoration.block_chunks_left.contains(h))
.chain(
restoration.manifest.state_hashes
.iter()
.filter(|h| !restoration.state_chunks_left.contains(h))
)
.map(|h| *h)
.collect();
Some(completed_chunks)
},
None => None,
}
}
fn chunk(&self, hash: H256) -> Option<Bytes> {
self.reader.read().as_ref().and_then(|r| r.chunk(hash).ok())
}
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
fn status(&self) -> RestorationStatus {
let mut cur_status = self.status.lock();
match *cur_status {
RestorationStatus::Initializing { ref mut chunks_done, .. } => {
*chunks_done = self.state_chunks.load(Ordering::SeqCst) as u32 +
self.block_chunks.load(Ordering::SeqCst) as u32;
}
RestorationStatus::Ongoing { ref mut state_chunks_done, ref mut block_chunks_done, .. } => {
*state_chunks_done = self.state_chunks.load(Ordering::SeqCst) as u32;
*block_chunks_done = self.block_chunks.load(Ordering::SeqCst) as u32;
},
_ => (),
}
cur_status.clone()
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
fn begin_restore(&self, manifest: ManifestData) {
2016-10-30 09:56:34 +01:00
if let Err(e) = self.io_channel.lock().send(ClientIoMessage::BeginRestoration(manifest)) {
trace!("Error sending snapshot service message: {:?}", e);
}
}
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
fn abort_restore(&self) {
trace!(target: "snapshot", "Aborting restore");
self.restoring_snapshot.store(false, Ordering::SeqCst);
*self.restoration.lock() = None;
*self.status.lock() = RestorationStatus::Inactive;
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
fn restore_state_chunk(&self, hash: H256, chunk: Bytes) {
2016-10-30 09:56:34 +01:00
if let Err(e) = self.io_channel.lock().send(ClientIoMessage::FeedStateChunk(hash, chunk)) {
trace!("Error sending snapshot service message: {:?}", e);
}
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
fn restore_block_chunk(&self, hash: H256, chunk: Bytes) {
2016-10-30 09:56:34 +01:00
if let Err(e) = self.io_channel.lock().send(ClientIoMessage::FeedBlockChunk(hash, chunk)) {
trace!("Error sending snapshot service message: {:?}", e);
}
Snapshot creation and restoration (#1679) * to_rlp takes self by-reference * clean up some derefs * out-of-order insertion for blockchain * implement block rebuilder without verification * group block chunk header into struct * block rebuilder does verification * integrate snapshot service with client service; flesh out implementation more * initial implementation of snapshot service * remove snapshottaker trait * snapshot writer trait with packed and loose implementations * write chunks using "snapshotwriter" in service * have snapshot taking use snapshotwriter * implement snapshot readers * back up client dbs when replacing * use snapshot reader in snapshot service * describe offset format * use new get_db_path in parity, allow some errors in service * blockchain formatting * implement parity snapshot * implement snapshot restore * force blocks to be submitted in order * fix bug loading block hashes in packed reader * fix seal field loading * fix uncle hash computation * fix a few bugs * store genesis state in db. reverse block chunk order in packed writer * allow out-of-order import for blocks * bring restoration types together * only snapshot the last 30000 blocks * restore into overlaydb instead of journaldb * commit version to database * use memorydbs and commit directly * fix trie test compilation * fix failing tests * sha3_null_rlp, not H256::zero * move overlaydb to ref_overlaydb, add new overlaydb without on-disk rc * port archivedb to new overlaydb * add deletion mode tests for overlaydb * use new overlaydb, check state root at end * share chain info between state and block snapshotting * create blocks snapshot using blockchain directly * allow snapshot from arbitrary block, remove panickers from snapshot creation * begin test framework * blockchain chunking test * implement stateproducer::tick * state snapshot test * create block and state chunks concurrently, better restoration informant * fix tests * add deletion mode tests for overlaydb * address comments * more tests * Fix up tests. * remove a few printlns * add a little more documentation to `commit` * fix tests * fix ref_overlaydb test names * snapshot command skeleton * revert ref_overlaydb renaming * reimplement snapshot commands * fix many errors * everything but inject * get ethcore compiling * get snapshot tests passing again * instrument snapshot commands again * fix fallout from other changes, mark snapshots as experimental * optimize injection patterns * do two injections * fix up tests * take snapshots from 1000 blocks efore * address minor comments * fix a few io crate related errors * clarify names about total difficulty [ci skip]
2016-08-05 17:00:46 +02:00
}
fn abort_snapshot(&self) {
if self.taking_snapshot.load(Ordering::SeqCst) {
trace!(target: "snapshot", "Aborting snapshot Snapshot under way");
self.progress.write().abort = true;
}
}
fn shutdown(&self) {
trace!(target: "snapshot", "Shut down SnapshotService");
self.abort_restore();
trace!(target: "snapshot", "Shut down SnapshotService - restore aborted");
self.abort_snapshot();
trace!(target: "snapshot", "Shut down SnapshotService - snapshot aborted");
}
}
impl<C: Send + Sync> Drop for Service<C> {
2016-09-06 15:41:56 +02:00
fn drop(&mut self) {
trace!(target: "shutdown", "Dropping Service");
2016-09-06 15:41:56 +02:00
self.abort_restore();
trace!(target: "shutdown", "Dropping Service - restore aborted");
self.abort_snapshot();
trace!(target: "shutdown", "Dropping Service - snapshot aborted");
2016-09-06 15:41:56 +02:00
}
}