Each Deploy

yarn codegen yarn build yarn deploy

Debug problems locally

Debugging can be difficult and mostly extremely time consuming as you need to wait hours/days/weeks for the new deploy to sync just to see if you actually fixed the problem or not.

The following steps will allow you to fork a current live deploy from the block before it started having problems and then with a local deploy (starting from this 'last synced block' so you wont have to wait for any sync) you can troubleshoot without the usual wait time.

1. Start a local graph node with the fork-base option set to: https://api.thegraph.com/subgraphs/id/, to allow you to fork a subgraph from the HostedService. Follow the steps here including the required dependencies: https://github.com/graphprotocol/graph-node#running-a-local-graph-node

Then: cargo run -p graph-node --release -- \ --postgres-url postgresql://USERNAME[:PASSWORD]@localhost:5432/graph-node \ --ethereum-rpc NETWORK_NAME:[CAPABILITIES]:URL \ --ipfs 127.0.0.1:5001 --fork-base https://api.thegraph.com/subgraphs/id/

ie: cargo run -p graph-node --release -- --postgres-url postgresql://postgres:MyPassword123@localhost:5432/graph-node --ethereum-rpc bsc:https://bsc-dataseed.binance.org/ --ipfs 127.0.0.1:5001 --fork-base https://api.thegraph.com/subgraphs/id/

2. Make changes to your subgraph to try fix the problem

3. Change the startBlock for each data source in subgraph.yaml to the problematic block

4. Deploy the forked subgraph locally graph deploy <subgraph-name> --debug-fork <subgraph-id> --ipfs http://localhost:5001 --node http://localhost:8020 <subgraph-name> should be consistent per deploy of the same project <subgraph-id> should change per each modified deploy An example: graph deploy spartan-protocol/pool-factory --debug-fork QmekrGD5Am9AKr3vapjrWrdfMNZrhcdZRGyUxRoV81pzpJ --ipfs http://localhost:5001 --node http://localhost:8020

5. Inspect logs, debug/troubleshoot. If the problem is fixed, change the startBlock for each data source in subgraph.yaml back to their previous values and re-deploy to hosted: yarn codegen yarn build yarn deploy

If the problem however is not fixed, start back at step #2