Subgraph saucerswap-v1 has indexing issues

[AlfredoG87]

Description

{
    "data": {
        "indexingStatusesForSubgraphName": [
            {
                "health": "failed",
                "nonFatalErrors": [],
                "fatalError": {
                    "message": "Mapping aborted at ~lib/@graphprotocol/graph-ts/chain/ethereum.ts, line 448, column 6, with message: Call reverted, probably because an `assert` or `require` in the contract failed, consider using `try_getPair` to handle this in the mapping.\twasm backtrace:\t    0: 0x1c43 - <unknown>!~lib/@graphprotocol/graph-ts/chain/ethereum/ethereum.SmartContract#call\t    1: 0x328a - <unknown>!src/types/templates/Pair/Factory/Factory#getPair\t    2: 0x3361 - <unknown>!src/mappings/pricing/findEthPerToken\t    3: 0x3760 - <unknown>!src/mappings/core/handleSync\t in handler `handleSync` at block #43665956 (f85dc43b434a8ca1620441f5923b14358148607f08d48c3547aa25e3fc8e6a61)",
                    "block": {
                        "number": "43665956",
                        "hash": "0xf85dc43b434a8ca1620441f5923b14358148607f08d48c3547aa25e3fc8e6a61"
                    },
                    "handler": null
                },
                "chains": [
                    {
                        "chainHeadBlock": {
                            "number": "65148267"
                        },
                        "latestBlock": {
                            "number": "43665956"
                        }
                    }
                ],
                "synced": false,
                "subgraph": "QmfAMQgq7VmDk3Z7ftwiLiHRUW6SndqxYCkGkZoTrSr27d"
            }
        ]
    }
}

Steps to reproduce

Run the following GraphQL query on the indexing endpoint for mainnet

https://mainnet-thegraph.swirldslabs.com/graphql

{
    indexingStatusesForSubgraphName(subgraphName: "saucerswap/saucerswap-v1") {
        health
        nonFatalErrors {
            message
            block {
                number
                hash
            }
            handler
            deterministic
        }
        fatalError {
            message
            block {
                number
                hash
            }
            handler
        }
        chains {
            chainHeadBlock {
                number
            }
            latestBlock {
                number
            }
        }
        synced
        subgraph
    }
}

Additional context

No response

Hedera network

mainnet

Version

v0.1.1

Operating system

None

[acuarica]

After further investigation, we solve this issue by clearing the call cache of The Graph.

Important

This issue happens in our current deployed version v0.29.0. It is fixed in The Graph v0.34.0 https://github.com/graphprotocol/graph-node/blob/master/NEWS.md#v0340. Even if this issue was solved, it is still a good idea to upgrade The Graph #87.

The issue is that a transient faulty eth_call result still gets cached, and subsequent eth_calls with the same arguments will return the original faulty result. See graphprotocol/graph-node#4756 for more info.

These are the steps we took to solve this problem, i.e., clear the cache and re-start the subgraph indexing.

• Check subgraph deployment status graphman --config config.toml info <deployment>

  $ graphman --config config.toml info Qme11uh1hK7zsxRRdszUJVTKgopuYuBxi2ZEvZbhaPgrHV
  name      | saucerswap/saucerswap-v1
  status    | current
  id        | Qme11uh1hK7zsxRRdszUJVTKgopuYuBxi2ZEvZbhaPgrHV
  shard     | primary
  active    | true
  chain     | mainnet
  node_id   | ---

• Clear Call cache graphman --config config.toml chain call-cache <chain> remove

  $ graphman --config config.toml chain call-cache mainnet remove
  The call cache has cleared

• Rewind deployment to given Block since we are not caching blocks, we need to include --force option graphman --config config.toml rewind --force <block-hash> <block-number> <deployment

  $ graphman --config config.toml rewind --force "0x488045e8e8ee45d33ee7209bb8459180c80017f2c8afe5482d4d56eb884f5fc2" "43665916" "Qme11uh1hK7zsxRRdszUJVTKgopuYuBxi2ZEvZbhaPgrHV"
  Pausing deployments
  
  Rewinding deployments
    ... rewound Qme11uh1hK7zsxRRdszUJVTKgopuYuBxi2ZEvZbhaPgrHV[14]
  Resuming deployments

• Re-assign deployment so an indexer resumes work graphman --config config.toml reassign <deploymentId> index_node

  $ graphman --config config.toml reassign Qme11uh1hK7zsxRRdszUJVTKgopuYuBxi2ZEvZbhaPgrHV index_node
  assigning Qme11uh1hK7zsxRRdszUJVTKgopuYuBxi2ZEvZbhaPgrHV[14] to index_node

• Verify the subgraph deployment is assigned a node

  $ graphman --config config.toml info Qme11uh1hK7zsxRRdszUJVTKgopuYuBxi2ZEvZbhaPgrHV
  name      | saucerswap/saucerswap-v1
  status    | current
  id        | Qme11uh1hK7zsxRRdszUJVTKgopuYuBxi2ZEvZbhaPgrHV
  shard     | primary
  active    | true
  chain     | mainnet
  node_id   | index_node