Allow closing allocations when maxBlock is set in manifest

[dwerner]

Previously, subgraphs continued scanning even when all the data sources had passed the end block.
This update finds the maximum end block across the data sources and stops the subgraph from scanning once the max_end_block is reached.

The agent won't be able to close allocations for those subgraphs because they aren't at chainhead.

related: graphprotocol/graph-node#5583

[dwerner]

Reproducing this with local-network:

1. create a subgraph with endBlock defined (1000 in my case)
2. deploy and allocate to it
3. mine blocks past the end block defined
4. verify that latestBlock in indexerStatus is remaining at 1000 while chainHeadBlock keeps on going
5. queue up an unallocate and wait for the action to be executed
6. observe this error below:

{
    "action": 3,
    "component": "AllocationManager",
    "error": {
        "cause": "POI not available for deployment at current epoch start block.\n              currentEpochStartBlock: 1210\n              deploymentStatus: [{\"subgraphDeployment\":{\"kind\":\"deployment-id\",\"value\":\"0x07e7cae9b5004024f70ad7d0f473b5b1867a7698fc871b9816e625a210d751fd\"},\"synced\":true,\"health\":\"healthy\",\"fatalError\":null,\"node\":\"default\",\"chains\":[{\"network\":\"hardhat\",\"latestBlock\":{\"number\":\"1000\",\"hash\":\"6a7425df7ae6b0e3cba250a85b1b370e49c05e1b08910732903731b8479af38c\"},\"chainHeadBlock\":{\"number\":\"1211\",\"hash\":\"d7487e29ab1280039235fc9d0faab92863cd85dc475a6660351fa20f44efa3c6\"},\"earliestBlock\":{\"number\":\"0\",\"hash\":\"0x0\"}}]}]",
        "code": "IE067",
        "explanation": "https://github.com/graphprotocol/indexer/blob/main/docs/errors.md#ie067"
    },
    "formatted_time": "2025-02-04 15:43:15:223",
    "hostname": "c829539864db",
    "level": 50,
    "level_text": "ERROR",
    "msg": "Failed to prepare tx call data",
    "name": "IndexerAgent",
    "pid": 173,
    "protocolNetwork": "eip155:1337",
    "time": 1738712595223
}

[dwerner]

https://github.com/graphprotocol/indexer/blob/main/packages/indexer-common/src/indexer-management/monitor.ts#L951

[dwerner]

In my fix here #1085

{
    "action": 4,
    "allocation": "0xcD0A322b8177228bEBe503E1e5BEb94d17CcC32F",
    "amountGRT": "0.005",
    "component": "AllocationManager",
    "deployment": {
        "bytes32": "0x07e7cae9b5004024f70ad7d0f473b5b1867a7698fc871b9816e625a210d751fd",
        "ipfsHash": "QmNsWZYpBmqXm4zEZRq5GpaB1BJ2J1hP8LgwW5ySYSrK3i"
    },
    "formatted_time": "2025-02-05 11:36:40:956",
    "hostname": "c829539864db",
    "indexer": "0xf4EF6650E48d099a4972ea5B414daB86e1998Bd3",
    "indexingRewards": {
        "hex": "0x00",
        "type": "BigNumber"
    },
    "level": 30,
    "level_text": "INFO",
    "msg": "Successfully closed allocation",
    "name": "IndexerAgent",
    "pid": 5703,
    "poi": "0xa0385267ee1d724b48dab571b4e8b73e22d8726adc043e5f831fa74b5309dbbe",
    "protocolNetwork": "eip155:1337",
    "time": 1738784200956,
    "transaction": "0x9eb5cc2e69d7a35d14390bbd5f1d229105cc42095dd3230be3b696745761a950"
}

I am now able to resolve the PoI from the block automatically.