openethereum/parity/snapshot.rs
s3krit 3fd58bdcbd
Beta 2.5.3 (#10776)
* ethcore/res: activate atlantis classic hf on block 8772000 (#10766)

* fix docker tags for publishing (#10741)

* fix: aura don't add `SystemTime::now()` (#10720)

This commit does the following:
- Prevent overflow in `verify_timestamp()` by not adding `now` to found faulty timestamp
- Use explicit `CheckedSystemTime::checked_add` to prevent potential consensus issues because SystemTime is platform
depedent
- remove `#[cfg(not(time_checked_add))]` conditional compilation

* Update version

* Treat empty account the same as non-exist accounts in EIP-1052 (#10775)

* DevP2p: Get node IP address and udp port from Socket, if not included in PING packet (#10705)

* get node IP address and udp port from Socket, if not included in PING packet

* prevent bootnodes from being added to host nodes

* code corrections

* code corrections

* code corrections

* code corrections

* docs

* code corrections

* code corrections

* Apply suggestions from code review

Co-Authored-By: David <dvdplm@gmail.com>

* Add a way to signal shutdown to snapshotting threads (#10744)

* Add a way to signal shutdown to snapshotting threads

* Pass Progress to fat_rlps() so we can abort from there too.

* Checking for abort in a single spot

* Remove nightly-only weak/strong counts

* fix warning

* Fix tests

* Add dummy impl to abort snapshots

* Add another dummy impl for TestSnapshotService

* Remove debugging code

* Return error instead of the odd Ok(())
Switch to AtomicU64

* revert .as_bytes() change

* fix build

* fix build maybe
2019-06-25 13:38:29 +00:00

295 lines
9.3 KiB
Rust

// Copyright 2015-2019 Parity Technologies (UK) Ltd.
// This file is part of Parity Ethereum.
// Parity Ethereum is free software: you can redistribute it and/or modify
// 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,
// 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 and restoration commands.
use std::time::Duration;
use std::path::{Path, PathBuf};
use std::sync::Arc;
use hash::keccak;
use ethcore::snapshot::{Progress, RestorationStatus, SnapshotConfiguration, SnapshotService as SS};
use ethcore::snapshot::io::{SnapshotReader, PackedReader, PackedWriter};
use ethcore::snapshot::service::Service as SnapshotService;
use ethcore::client::{Mode, DatabaseCompactionProfile, VMType};
use ethcore::miner::Miner;
use ethcore_service::ClientService;
use types::ids::BlockId;
use cache::CacheConfig;
use params::{SpecType, Pruning, Switch, tracing_switch_to_bool, fatdb_switch_to_bool};
use helpers::{to_client_config, execute_upgrades};
use dir::Directories;
use user_defaults::UserDefaults;
use ethcore_private_tx;
use db;
/// Kinds of snapshot commands.
#[derive(Debug, PartialEq, Clone, Copy)]
pub enum Kind {
/// Take a snapshot.
Take,
/// Restore a snapshot.
Restore
}
/// Command for snapshot creation or restoration.
#[derive(Debug, PartialEq)]
pub struct SnapshotCommand {
pub cache_config: CacheConfig,
pub dirs: Directories,
pub spec: SpecType,
pub pruning: Pruning,
pub pruning_history: u64,
pub pruning_memory: usize,
pub tracing: Switch,
pub fat_db: Switch,
pub compaction: DatabaseCompactionProfile,
pub file_path: Option<String>,
pub kind: Kind,
pub block_at: BlockId,
pub max_round_blocks_to_import: usize,
pub snapshot_conf: SnapshotConfiguration,
}
// helper for reading chunks from arbitrary reader and feeding them into the
// service.
fn restore_using<R: SnapshotReader>(snapshot: Arc<SnapshotService>, reader: &R, recover: bool) -> Result<(), String> {
let manifest = reader.manifest();
info!("Restoring to block #{} (0x{:?})", manifest.block_number, manifest.block_hash);
snapshot.init_restore(manifest.clone(), recover).map_err(|e| {
format!("Failed to begin restoration: {}", e)
})?;
let (num_state, num_blocks) = (manifest.state_hashes.len(), manifest.block_hashes.len());
let informant_handle = snapshot.clone();
::std::thread::spawn(move || {
while let RestorationStatus::Ongoing { state_chunks_done, block_chunks_done, .. } = informant_handle.status() {
info!("Processed {}/{} state chunks and {}/{} block chunks.",
state_chunks_done, num_state, block_chunks_done, num_blocks);
::std::thread::sleep(Duration::from_secs(5));
}
});
info!("Restoring state");
for &state_hash in &manifest.state_hashes {
if snapshot.status() == RestorationStatus::Failed {
return Err("Restoration failed".into());
}
let chunk = reader.chunk(state_hash)
.map_err(|e| format!("Encountered error while reading chunk {:?}: {}", state_hash, e))?;
let hash = keccak(&chunk);
if hash != state_hash {
return Err(format!("Mismatched chunk hash. Expected {:?}, got {:?}", state_hash, hash));
}
snapshot.feed_state_chunk(state_hash, &chunk);
}
info!("Restoring blocks");
for &block_hash in &manifest.block_hashes {
if snapshot.status() == RestorationStatus::Failed {
return Err("Restoration failed".into());
}
let chunk = reader.chunk(block_hash)
.map_err(|e| format!("Encountered error while reading chunk {:?}: {}", block_hash, e))?;
let hash = keccak(&chunk);
if hash != block_hash {
return Err(format!("Mismatched chunk hash. Expected {:?}, got {:?}", block_hash, hash));
}
snapshot.feed_block_chunk(block_hash, &chunk);
}
match snapshot.status() {
RestorationStatus::Ongoing { .. } => Err("Snapshot file is incomplete and missing chunks.".into()),
RestorationStatus::Initializing { .. } => Err("Snapshot restoration is still initializing.".into()),
RestorationStatus::Failed => Err("Snapshot restoration failed.".into()),
RestorationStatus::Inactive => {
info!("Restoration complete.");
Ok(())
}
}
}
impl SnapshotCommand {
// shared portion of snapshot commands: start the client service
fn start_service(self) -> Result<ClientService, String> {
// load spec file
let spec = self.spec.spec(&self.dirs.cache)?;
// load genesis hash
let genesis_hash = spec.genesis_header().hash();
// database paths
let db_dirs = self.dirs.database(genesis_hash, None, spec.data_dir.clone());
// user defaults path
let user_defaults_path = db_dirs.user_defaults_path();
// load user defaults
let user_defaults = UserDefaults::load(&user_defaults_path)?;
// select pruning algorithm
let algorithm = self.pruning.to_algorithm(&user_defaults);
// check if tracing is on
let tracing = tracing_switch_to_bool(self.tracing, &user_defaults)?;
// check if fatdb is on
let fat_db = fatdb_switch_to_bool(self.fat_db, &user_defaults, algorithm)?;
// prepare client and snapshot paths.
let client_path = db_dirs.client_path(algorithm);
let snapshot_path = db_dirs.snapshot_path();
// execute upgrades
execute_upgrades(&self.dirs.base, &db_dirs, algorithm, &self.compaction)?;
// prepare client config
let mut client_config = to_client_config(
&self.cache_config,
spec.name.to_lowercase(),
Mode::Active,
tracing,
fat_db,
self.compaction,
VMType::default(),
"".into(),
algorithm,
self.pruning_history,
self.pruning_memory,
true,
self.max_round_blocks_to_import,
);
client_config.snapshot = self.snapshot_conf;
let restoration_db_handler = db::restoration_db_handler(&client_path, &client_config);
let client_db = restoration_db_handler.open(&client_path)
.map_err(|e| format!("Failed to open database {:?}", e))?;
let service = ClientService::start(
client_config,
&spec,
client_db,
&snapshot_path,
restoration_db_handler,
&self.dirs.ipc_path(),
// TODO [ToDr] don't use test miner here
// (actually don't require miner at all)
Arc::new(Miner::new_for_tests(&spec, None)),
Arc::new(ethcore_private_tx::DummySigner),
Box::new(ethcore_private_tx::NoopEncryptor),
Default::default(),
Default::default(),
).map_err(|e| format!("Client service error: {:?}", e))?;
Ok(service)
}
/// restore from a snapshot
pub fn restore(self) -> Result<(), String> {
let file = self.file_path.clone();
let service = self.start_service()?;
warn!("Snapshot restoration is experimental and the format may be subject to change.");
warn!("On encountering an unexpected error, please ensure that you have a recent snapshot.");
let snapshot = service.snapshot_service();
if let Some(file) = file {
info!("Attempting to restore from snapshot at '{}'", file);
let reader = PackedReader::new(Path::new(&file))
.map_err(|e| format!("Couldn't open snapshot file: {}", e))
.and_then(|x| x.ok_or("Snapshot file has invalid format.".into()));
let reader = reader?;
restore_using(snapshot, &reader, true)?;
} else {
info!("Attempting to restore from local snapshot.");
// attempting restoration with recovery will lead to deadlock
// as we currently hold a read lock on the service's reader.
match *snapshot.reader() {
Some(ref reader) => restore_using(snapshot.clone(), reader, false)?,
None => return Err("No local snapshot found.".into()),
}
}
Ok(())
}
/// Take a snapshot from the head of the chain.
pub fn take_snapshot(self) -> Result<(), String> {
let file_path = self.file_path.clone().ok_or("No file path provided.".to_owned())?;
let file_path: PathBuf = file_path.into();
let block_at = self.block_at;
let service = self.start_service()?;
warn!("Snapshots are currently experimental. File formats may be subject to change.");
let writer = PackedWriter::new(&file_path)
.map_err(|e| format!("Failed to open snapshot writer: {}", e))?;
let progress = Arc::new(Progress::default());
let p = progress.clone();
let informant_handle = ::std::thread::spawn(move || {
::std::thread::sleep(Duration::from_secs(5));
let mut last_size = 0;
while !p.done() {
let cur_size = p.size();
if cur_size != last_size {
last_size = cur_size;
let bytes = ::informant::format_bytes(cur_size as usize);
info!("Snapshot: {} accounts {} blocks {}", p.accounts(), p.blocks(), bytes);
}
::std::thread::sleep(Duration::from_secs(5));
}
});
if let Err(e) = service.client().take_snapshot(writer, block_at, &*progress) {
let _ = ::std::fs::remove_file(&file_path);
return Err(format!("Encountered fatal error while creating snapshot: {}", e));
}
info!("snapshot creation complete");
assert!(progress.done());
informant_handle.join().map_err(|_| "failed to join logger thread")?;
Ok(())
}
}
/// Execute this snapshot command.
pub fn execute(cmd: SnapshotCommand) -> Result<String, String> {
match cmd.kind {
Kind::Take => cmd.take_snapshot()?,
Kind::Restore => cmd.restore()?,
}
Ok(String::new())
}