pystarport: A Python repository from GAtom22

pystarport is like a cosmos starport without the scaffolding feature. it's mainly used for development and testing. It's developed for the Crypto.org Chain, but it can also be used for any cosmos-sdk based projects.

Configuration

a typical configuration for a devnet is like this:

chainmaind:
  cmd: chain-maind  # chain binary to use, optional
  validators:  # genesis validators
    - coins: 10cro
      staked: 10cro
    - coins: 10cro
      staked: 10cro
  accounts:  # genesis accounts
    - name: community
      coins: 100cro
    - name: ecosystem
      coins: 200cro
    - name: reserve
      coins: 200cro
      vesting_coins: 100cro # if not provided, the all of coins will be vested
      vesting: "1d"
    - name: launch
      coins: 100cro
  genesis:  # patch genesis states
   app_state:
     staking:
       params:
         unbonding_time: "10s"

The validators section defines how many nodes to run, for each node, a home directory is initialized in data/node{i}, and a validator account with specified coins is created.

The accounts defines other non-validator accounts, they are created in node0's keyring.

In the genesis section you can override any genesis configuration with the same json path.

Usage

NAME
    pystarport serve - prepare and start a devnet from scatch

SYNOPSIS
    pystarport serve <flags>

DESCRIPTION
    prepare and start a devnet from scatch

FLAGS
    --data=DATA
        Type: str
        Default: './data'
        path to the root data directory
    --config=CONFIG
        Type: str
        Default: './config.yaml'
        path to the configuration file
    --base_port=BASE_PORT
        Type: int
        Default: 26650
        the base port to use, the service ports of different nodes are calculated based on this
    --cmd=CMD
        Type: str
        Default: 'chain-maind'
        the chain binary to use

Port rules

The rules to calculate service ports based on base port is defined in the ports.py module.

For example, with default base port 26650, the url of api servers of the nodes would be:

Node0: http://127.0.0.1:26654
Node1: http://127.0.0.1:26664

The swagger doc of node0 is http://127.0.0.1:26654/swagger/

The default rpc port used by chain-maind is 26657, that's the default node0's rpc port, so you can use chain-maind without change to access node0's rpc.

Supervisor

pystarport embeded a supervisor to manage processes of multiple nodes, you can use pystarport supervisorctl to manage the processes:

$ pystarport supervisorctl status
node0                            RUNNING   pid 35210, uptime 0:00:29
node1                            RUNNING   pid 35211, uptime 0:00:29
$ pystarport supervisorctl help

default commands (type help <topic>):
=====================================
add    exit      open  reload  restart   start   tail
avail  fg        pid   remove  shutdown  status  update
clear  maintail  quit  reread  signal    stop    version

Or enter an interactive shell:

$ pystarport supervisorctl
node0                            RUNNING   pid 35210, uptime 0:01:53
node1                            RUNNING   pid 35211, uptime 0:01:53
supervisor>

Cli

After started the chain, you can use chain-maind cli directly, there are also some wrapper commands provided by pystarport cli. It understands the directory structure and port rules, also assuming keyring-backend=test, and there are shortcuts for commonly used commands, so arguments are shorter.

$ pystarport cli - --help
...

Transaction Bot

A simple transaction bot that works for cluster created by pystarport as well as a local node

Copy and modify bot.yaml.sample to bot.yaml with your desired bot configurations.

If you are running on a pystarport created cluster:

Make sure you have provide the node for each job in the bot.yaml
Run the command

$ pystarport bot --chain-id=[cluster_chain_id] - start

If you are running on a local node

$ pstarport bot --node_rpc=tcp://127.0.0.1:26657 --data=/path/to/your/local/node/home/ - start

docker-compose

When used with docker-compose or multiple machines, you need to config hostnames, and you probabely want to use a same base_port since you don't have port conflicts, you can config the validators like this:

validators:
  - coins: 10cro
    staked: 10cro
    base_port: 26650
    hostname: node0
  - coins: 10cro
    staked: 10cro
    base_port: 26650
    hostname: node1

pystarport init --gen_compose_file will also generate a docker-compose.yml file for you.

IBC

It can setup multiple devnets at once, and connect them with ibc relayer.

ibc-0:
  validators:
    - coins: 10cro
      staked: 10cro
      base_port: 26650
    - coins: 10cro
      staked: 10cro
  accounts:
    - name: relayer
      coins: 100cro
  genesis:
    app_state:
      transfer:
        params:
          receive_enabled: true
          send_enabled: true
ibc-1:
  validators:
    - coins: 10cro
      staked: 10cro
      base_port: 26750
    - coins: 10cro
      staked: 10cro
      base_port: 26760
  accounts:
    - name: relayer
      coins: 100cro
  genesis:
    app_state:
      transfer:
        params:
          receive_enabled: true
          send_enabled: true
relayer: # refer here (https://hermes.informal.systems/config.html) for more configs
  global:
    strategy: 'packets'
    log_level: 'info'
  chains:
    - id: "ibc-0" # id is needed to match chain id
      trusting_period: "20s"
    - id: "ibc-1" # id is needed to match chain id
      trusting_period: "20s"

With following commands to setup ibc, you are ready to play with ibc functionalities:

# spawn the devnets
pystarport serve --config ibc.yaml
# setup ibc channel
hermes -c data/relayer.toml create channel ibc-0 ibc-1 --port-a transfer --port-b transfer
# start relayer process
supervisorctl -c data/tasks.ini start relayer-demo

Development

Set up development environment

More about poetry.

poetry install

Recommended VS Code workspace settings

Remember to run poetry env info after poetry install and update this python.defaultInterpreterPath

{
    "python.formatting.provider": "black",
    // after running `poetry env info`, change python.defaultInterpreterPath to Path under Virtualenv
    "python.defaultInterpreterPath": "~/Library/Caches/pypoetry/virtualenvs/to_be_updated",
    "editor.formatOnSave": true,
    "python.linting.flake8Enabled": true,
    "python.formatting.blackArgs": [
        "--line-length=88"
    ],
    "python.sortImports.args": [
        "--profile",
        "black"
    ],
    "[python]": {
        "editor.codeActionsOnSave": {
            "source.organizeImports": true
        }
    }
}

Test

Install jsonnet

More about jsonnet.

make test

FAQ

Have this error on Mac: `AF_UNIX path too long´

Set environment variable TMPPATH=/tmp to shorten the unix domain socket path.

GAtom22/pystarport