CIC STAFF CLIENT

Services installer temporarily for internal use by GE.

Dependencies

The os-level dependencies below must be met both at install and run time.

The version numbers are the version numbers used at implmentation time. It may very well work with earlier versions of the components, as long as they are not too old. An internet connection will be needed if the python dependencies cannot be resolved locally, either through a local repository or an existing package cache.

• systemd (249)
• gcc (11.1.0)
• git (2.33.0)
• python (>= 3.9)
• pip (20.3.4)
• sqlite (3.36.0)

For the optional bloxberg node build (INSTALL_EVM=bloxberg), additionally these dependencies must be met, aswell as a working internet connection:

• rustup (1.24.3)
• clang (12.0.1)
• cmake (3.21.2)

Installation settings

The examples below assume working directory of the cic-staff-installer repository root.

cic-stack docker-compose cluster settings

To use against the cic-stack docker-compose local cluster:

export RPC_PROVIDER=http://localhost:63545
export CIC_ROOT_URL=file://`pwd`/var

If you want to select python packages from a specific repository only, also add:

export PIP_INDEX_URL=<url>
export PIP_EXTRA_INDEX_URL=<url>

Bloxberg

If you want to build the bloxberg mode executable (be warned, that's a long wait), add:

export INSTALL_EVM=bloxberg

If you wish, you can use an existing openethereum executable instead. Make sure <path_to_binary> --version shows 2.7.2, and then set:

export OPENETHEREUM_PATH=<path_to_binary>

Installation

To proceed with the installation, enter:

bash setup.sh

During the installation you will be prompted to enter your name, email and as password for the gnupg setup.

The gnupg key will both be used to authenticate using HTTP HOBA when necessary, aswell as encrypt local cached content.

Running the services

systemctl --user start cic-cache-tracker
systemctl --user start cic-cache-server

Verify that they are running

systemctl --user status cic-cache-tracker
systemctl --user status cic-cache-server

The bloxberg node, if installed, runs in the same manner:

systemctl --user start bloxberg
systemctl --user status bloxberg

Using clicada

It should now be possible to run clicada without any extra settings needed.

Please refer to the documentation on clicada for details on how to use the tool.

Files and directories

The installation produces a number of files in the user home directory, some of which may be edited directly to change behavior of the program.

All paths relative to $HOME

location	description	editable
.config/cic/cache/*.ini	Configuration file(s) for the cic-cache-* services	yes
.config/cic/cache/*.ini	Configuration file(s) for the cic-cache-* services	yes
.config/cic/clicada/*.ini	Configuration file(s) for the clicada tool	yes
.config/cic/staff-client/key_fingerprint	gnupg key fingerprint for key used by clicada for authentication	no
.config/cic/staff-client/user.asc	gnupg public key used by cicada for authentication	no
.config/cic/staff-client/.gnupg	gnupg homedir used by cicada for authentication	no
.config/systemd/user/cic-cache-*.service	systemd user service definition file for cic-cache-* services	yes, with systemctl --user edit <service>
.config/systemd/user/bloxberg.service	systemd user service definition file for bloxberg	yes, with systemctl --user edit <service>
.config/environment.d/01-cic-cache-*.conf	environment variables for systemd user services	yes
.local/share/cic/.gnupg	gnupg homedir for holding trust keys for global cic configurations	no
.local/share/cic/clicada/.secret	A gnupg encrypted symmetric secret used to encrypt local cached content	no
.local/share/io.parity.ethereum/bloxberg	Bloxberg chain data	no
.config/io.parity.ethereum/bloxberg	Bloxberg configurations and chain data	bootnode.toml and bootnode.txt only

Installing as a different user

You may want to create a dedicated user for the installation, so as to not pollute your regular user data directories.

Since the services are run using systemd, a simple su or sudo will not be sufficient in this case.

Perhaps the simplest solution is to launch a new login shell within the systemd vm using the following command:

machinectl login

Another alternative can be to open an ssh session.