Skip to main content

Run

To get a full list of commands and options, run the following command:

./zig-out/bin/sig --help

The binary located at ./zig-out/bin/sig will be abbreviated to sig in the rest of this document.

Validator​

Startup the validator with the following command:

sig validator

Note: this command is modified as more validator components are completed.

Identity​

Sig stores its private key in ~/.sig/identity.key. On its first run, Sig will automatically generate a key if no key exists. To see the public key, use the identity subcommand.

sig identity

Gossip​

To run Sig as a Solana gossip client, use the gossip subcommand.

  • Specify entrypoints to connect to a cluster using -e or use pre-defined entrypoints with the -n option, followed by the network you are connecting to.
  • Optionally use -p to specify a custom listening port (default is 8001).

For more info about gossip, see the readme.

sig gossip -n <NETWORK>

where <NETWORK> is either mainnet, devnet, or testnet.

or

sig gossip -p <PORT> --entrypoint <IP>:<PORT>

AccountsDB​

We also support a few accounts-db related commands.

Download a snapshot​

You can download a snapshot from the network:

sig download_snapshot \
    # where to save snapshot
    -s test_data/tmp \
    # use default testnet entrypoints for gossip
    -n testnet \
    # pubkeys of validators who you trust
    --trusted-validator 3gxDv5XbkkXUiqKiqt5WbsfGN7i9GHB1xMWBqhg4UDzj \
    # minimum MB/s speed when downloading snapshot
    --min-snapshot-download-speed 50

Note: we dont support downloading genesis files, however they can be found in data/genesis-files/ in the repo.

Validate a snapshot​

You can also validate an existing snapshot:

sig snapshot-validate \
    # path to snapshot
    -s test_data/tmp \
    # validate against the genesis file for testnet
    -n testnet \
    # stream the accounts out of the snapshot (see the geyser readme to know how to read from the stream)
    --enable-geyser \
    # save the index generated for faster loads in the future (use --fastload in future runs)
    --save-index \
    # preallocate some accounts to increase the speed of the snapshot validation
    --accounts-per-file-estimate 500

Note: if you are running out-of-memory when loading, you can either reduce the preallocations using a smaller value for --accounts-per-file-estimate or use --use-disk-index to use disk memory instead of RAM.

Note: you can also define a specifc genesis file to validate against using -g.

Collect Shreds​

You can also connect to the network and collect shreds:

sig shred-network \
    # collect shreds from testnet
    -n testnet \
    # dont re-transmit the shreds in the turbine-tree (only recv)
    --no-retransmit