stack-orchestrator/stack_orchestrator/data/stacks/graph-node

Graph Node

Setup

Clone required repositories:

laconic-so --stack graph-node setup-repositories --pull

Checkout to a non-default branch in the cloned repos if required:

# Default repo base dir
cd ~/cerc

# Example
cd graph-node
git checkout <your-branch> && git pull

# Remove the corresponding docker image if it already exists
docker image rm cerc/graph-node:local
# Remove any dangling images
docker image prune

Build the container images:

laconic-so --stack graph-node build-containers

Create a deployment

Initialize deployment and create "spec" file:

laconic-so --stack graph-node deploy init --output graph-node-spec.yml

We need to assign fixed ports: 8000 for subgraph GQL endpoint, 8020 for subgraph deployment and 5001 for IPFS. The values can be customized by editing the "spec" file generated by laconic-so deploy init.

$ cat graph-node-spec.yml
stack: graph-node
network:
  ports:
    graph-node:
      - '8000:8000'
      - '8001'
      - '8020:8020'
      - '8030'
      - '8040'
    ipfs:
      - '8080'
      - '4001'
      - '5001:5001'
...

Create deployment:

laconic-so --stack graph-node deploy create --spec-file graph-node-spec.yml --deployment-dir graph-node-deployment

Start the stack

Update config.env file inside deployment directory with the following values before starting the stack:

# Set ETH RPC endpoint the graph node will use

# Host and port of the ETH RPC endpoint to check before starting graph-node
export ETH_RPC_HOST=
export ETH_RPC_PORT=

# The etherum network(s) graph-node will connect to
# Set this to a space-separated list of the networks where each entry has the form NAME:URL
export ETH_NETWORKS=

# Optional:

# Timeout for ETH RPC requests in seconds (default: 180s)
export GRAPH_ETHEREUM_JSON_RPC_TIMEOUT=

# Number of times to retry ETH RPC requests (default: 10)
export GRAPH_ETHEREUM_REQUEST_RETRIES=

# Maximum number of blocks to scan for triggers in each request (default: 2000)
export GRAPH_ETHEREUM_MAX_BLOCK_RANGE_SIZE=

# Maximum number of concurrent requests made against Ethereum for requesting transaction receipts during block ingestion (default: 1000)
export GRAPH_ETHEREUM_BLOCK_INGESTOR_MAX_CONCURRENT_JSON_RPC_CALLS_FOR_TXN_RECEIPTS=

# Ref: https://git.vdb.to/cerc-io/graph-node/src/branch/master/docs/environment-variables.md

Example config.env file:

export ETH_RPC_HOST=filecoin.chainup.net
export ETH_RPC_PORT=443

export ETH_NETWORKS=filecoin:https://filecoin.chainup.net/rpc/v1

export GRAPH_ETHEREUM_JSON_RPC_TIMEOUT=360
export GRAPH_ETHEREUM_REQUEST_RETRIES=5
export GRAPH_ETHEREUM_MAX_BLOCK_RANGE_SIZE=50

Deploy the stack:

laconic-so deployment --dir graph-node-deployment start

# Note: Remove any existing volumes in the cluster for a fresh start

After all services have started, follow graph-node logs:

laconic-so deployment --dir graph-node-deployment logs -f graph-node

Subgraphs can now be deployed to the graph-node. Follow Deploy the Subgraph section in graph-node docs for an existing subgraph.

Set environment variables

• The graph-node environment variable ETHEREUM_REORG_THRESHOLD can be set in the deployment compose file

  $ cat graph-node-deployment/compose/docker-compose-graph-node.yml
  services:
    graph-node:
      image: cerc/graph-node:local
      ...
      environment:
        ...
        GRAPH_LOG: debug
        ETHEREUM_REORG_THRESHOLD: 3
  

  Change ETHEREUM_REORG_THRESHOLD to desired value

  • To restart graph-node with updated values
    • Stop the stack first

      laconic-so deployment --dir graph-node-deployment stop
      

    • Start the stack again

      laconic-so deployment --dir graph-node-deployment start
      

  • To check if environment variable has been updated in graph-node container

    $ laconic-so deployment --dir graph-node-deployment exec graph-node bash
    root@dc4d3abe1615:/# echo $ETHEREUM_REORG_THRESHOLD
    16
    

Clean up

Stop all the services running in background run:

laconic-so deployment --dir graph-node-deployment stop

Clear volumes created by this stack:

laconic-so deployment --dir graph-node-deployment stop --delete-volumes