[General] Run multiple local-nodes in different machines

[aulnes]

Environment

Testnet

L2 block number

No response

Provide a brief description of the functionality you're trying to implement and the issue you are running into.

I have build a local-node image from zksync-era main brach, and use the settings from matter-labs/local-setup to run the docker compose. I also use the test tools in archived repo matter-labs/era-tutorial-examples, and everything went ok.
Although I have had a running local-node(as well as a local chain), I have following questions:

Is it possible to deploy another local-node on another machine and interact with my local chain?
Just like a zksync-node read the mainnet node and get the chain states.

Repo Link (Optional)

https://github.com/matter-labs/local-setup

Additional Details

I use the zk tools to build a local-node from zksync-era, and I add the explorer and other staff(copy from matter-labs/local-setup/zk-chains-docker-compose.yml) in the original docker-compose.yml file. The local environment running normally, and the contracts can be deployed successfully.
I have not found any docs or specs introducing how to run multiple local-nodes in different machines, are these local-nodes totally independent?

[gleenai[bot]]

Running multiple local ZKsync nodes on different machines and having them interact with each other like a network is not directly supported by the standard matter-labs/local-setup. Each local node setup typically operates independently, primarily designed for isolated testing and development purposes.

However, if you're looking to simulate a more network-like environment where multiple nodes interact, you might need to customize the networking and synchronization aspects between these nodes manually. This could involve configuring the nodes to communicate over a shared network and ensuring they are synchronized to maintain the same chain state.

For detailed guidance on setting up such an environment, you might need to delve deeper into network configuration and possibly modify the Docker compose files to allow for inter-node communication. Unfortunately, there are no specific documents or specs provided in the current ZKsync documentation that describe running multiple local nodes in different machines in a networked manner.

For further assistance, you might consider reaching out directly to the ZKsync developer community or checking for updates in the ZKsync GitHub repositories for any new tools or features that might support this functionality in the future.

Your feedback helps us serve you better. If this fully addresses your question, please give it a thumbs up. If it doesn't, please give it a thumbs down. This will inform the community and team so we can improve our support.

[bxpana]

I don't see why it wouldn't be possible if you were using 2 different RPC URLs on the same network. What are you trying to do exactly so I can see if what you want is possible right now.

[aulnes]

Uploaded my settings.

[bxpana]

Appreciate you adding the details. I'll double check and let you know.

[bxpana]

Currently we don't have a laid out process for how to do this and it's something that we are working on. For now you can continue to test things out to see if you can get things working and in the future we'll be able to provide clearer details on how to do this.

[aulnes]

What I have done:

1. built a local-node image from zksync-era main branch, using zk tools like zk docker build local-node --custom-tag "mylocal" .
2. Running the local-node under the matter-labs/local-setup/docker-compose.yml settings, and I added some explorer docker images in the yml file to see the block(copy from matter-labs/local-setup/zk-chains-docker-compose.yml).
3. Deploy contracts on the local running chain, use matter-labs/era-tutorial-examples/local-setup-testing

What I am trying to do:

1. Running another "mylocal2" local-node(or external-node) on a different machine.
2. Let "mylocal" and "mylocal2" in same network, and they are on the same chain. It is like a zksync-node read and recovery the chain history of main node.
3. In short, let "mylocal" and "mylocal2" be decentralized nodes of one chain.

Below is my setting file docker-compose.yml:

services:
  reth:
    restart: always
    image: "ghcr.io/paradigmxyz/reth:v0.2.0-beta.2"
    volumes:
      - type: bind
        source: ./reth_chaindata
        target: /chaindata

    command: node --dev --datadir /rethdata --http --http.corsdomain "*" --http.addr 0.0.0.0 --http.port 8545  --dev.block-time 300ms --chain /chaindata/reth_config
    environment:
      - RUST_LOG=warn
    ports:
      - 127.0.0.1:15045:8545

  postgres:
    image: "postgres:14"
    # To create necessary databases for block explorer.
    volumes:
      - ./init.sql:/docker-entrypoint-initdb.d/init.sql
    command: postgres -c 'max_connections=200'
    ports:
      - 127.0.0.1:15432:5432
    environment:
      # We bind only to 127.0.0.1, so setting insecure password is acceptable here
      - POSTGRES_PASSWORD=notsecurepassword

  pgadmin4:
    image: dcagatay/pwless-pgadmin4:latest
    ports:
      - 127.0.0.1:15430:80
    environment:
      POSTGRES_USER: postgres
      POSTGRES_PASSWORD: notsecurepassword
      POSTGRES_HOST: postgres
      POSTGRES_PORT: 5432
    restart: unless-stopped

  mylocal:
    stdin_open: true
    tty: true
    image: matterlabs/local-node:${INSTANCE_TYPE:-mylocal}
    healthcheck:
      test: curl --fail http://localhost:3071/health || exit 1
      interval: 10s
      timeout: 5s
      retries: 200
      start_period: 30s
    environment:
      - DATABASE_PROVER_URL=postgresql://postgres:notsecurepassword@postgres:5432/prover_local
      - DATABASE_URL=postgresql://postgres:notsecurepassword@postgres:5432/zksync_local
      - ETH_CLIENT_WEB3_URL=http://reth:8545
      - LEGACY_BRIDGE_TESTING=1
    ports:
      - 15100:3050 # JSON RPC HTTP port
      - 15101:3051 # JSON RPC WS port
    depends_on:
      - reth
      - postgres
    volumes:
      - shared_config:/etc/env/target
      - shared_tokens:/etc/tokens
    extra_hosts:
      - host.docker.internal:host-gateway

  # L2 explorer: data-fetcher
  data-fetcher-main:
    platform: linux/amd64
    image: "matterlabs/block-explorer-data-fetcher:latest"
    environment:
      - PORT=3040
      - LOG_LEVEL=verbose
      - BLOCKCHAIN_RPC_URL=http://mylocal:3050
    ports:
      - 127.0.0.1:15104:3040
    restart: unless-stopped

  # L2 explorer: worker
  worker-main:
    platform: linux/amd64
    image: "matterlabs/block-explorer-worker:latest"
    environment:
      - PORT=3001
      - LOG_LEVEL=verbose
      - DATABASE_HOST=postgres
      - DATABASE_USER=postgres
      - DATABASE_PASSWORD=notsecurepassword
      - DATABASE_NAME=block-explorer
      - BLOCKCHAIN_RPC_URL=http://mylocal:3050
      - DATA_FETCHER_URL=http://data-fetcher-main:3040
      - BATCHES_PROCESSING_POLLING_INTERVAL=1000
    ports:
      - 127.0.0.1:15103:3001
    restart: unless-stopped

  # L2 explorer: API
  api-main:
    platform: linux/amd64
    image: "matterlabs/block-explorer-api:latest"
    environment:
      - PORT=3020
      - METRICS_PORT=3005
      - LOG_LEVEL=verbose
      - DATABASE_URL=postgresql://postgres:notsecurepassword@postgres:5432/block-explorer
    ports:
      - 127.0.0.1:3020:3020 # This is super tricky thing, as the explorer app must depends on api server running on port 3020. Should change the explorer app to have a better solution.
      - 127.0.0.1:15105:3005
    depends_on:
      - worker-main
    restart: unless-stopped

  # L2 explorer: APP
  app:
    platform: linux/amd64
    image: "matterlabs/block-explorer-app:latest"
    environment:
      - NODE_ENV=development
    ports:
      - 15005:3010
    depends_on:
      - api-main
    restart: unless-stopped

  # L1 Blockscout explorer: redis
  redis-db:
    image: 'redis:alpine'
    command: redis-server

  # L1 Blockscout explorer: backend
  backend:
    depends_on:
      - postgres
      - redis-db
    image: blockscout/blockscout:latest
    restart: always
    stop_grace_period: 5m
    command: sh -c "bin/blockscout eval \"Elixir.Explorer.ReleaseTasks.create_and_migrate()\" && bin/blockscout start"
    env_file:
      - blockscout/common-blockscout.env
    links:
      - postgres:database
    environment:
      - ETHEREUM_JSONRPC_VARIANT=geth
      - DATABASE_URL=postgresql://postgres:notsecurepassword@postgres:5432/blockscout
      - ETHEREUM_JSONRPC_HTTP_URL=http://reth:8545
      - ETHEREUM_JSONRPC_TRACE_URL=http://reth:8545

  # L1 Blockscout explorer: frontend
  frontend:
    depends_on:
      - backend
    image: ghcr.io/blockscout/frontend:latest
    platform: linux/amd64
    restart: always
    env_file:
      - blockscout/common-frontend.env

  stats:
    depends_on:
      - backend
    image: ghcr.io/blockscout/stats:latest
    platform: linux/amd64
    restart: always
    env_file:
      - blockscout/common-stats.env
    environment:
      - STATS__DB_URL=postgresql://postgres:notsecurepassword@postgres:5432/blockscout_stats
      - STATS__BLOCKSCOUT_DB_URL=postgresql://postgres:notsecurepassword@postgres:5432/blockscout
      - STATS__CREATE_DATABASE=true
      - STATS__RUN_MIGRATIONS=true

  proxy:
    depends_on:
      - backend
      - frontend
      - stats
    image: nginx
    volumes:
      - "./blockscout/proxy:/etc/nginx/templates"
    environment:
      BACK_PROXY_PASS: http://backend:4000
      FRONT_PROXY_PASS: http://frontend:3000
    ports:
      - target: 80
        published: 15001
      - target: 8080
        published: 15002
      - target: 8081
        published: 15003

  sig-provider:
    image: ghcr.io/blockscout/sig-provider:latest
    platform: linux/amd64
    restart: always

  hyperexplorer:
    depends_on:
      mylocal:
        condition: service_healthy
    image: ghcr.io/mm-zk/zksync_tools:latest
    ports:
      - 15000:5000
    volumes:
      - ./hyperexplorer.json:/app/operator/config.json

volumes:
  reth:
  shared_config:
  shared_tokens:

[aulnes]

Please let me know if you need more infos