[Bug] Rewinding a specific sgd rewinds another instance of the same deployment, causing data corruption

[madumas]

Bug report

For a duplicated subgraph, say sgdXXX and sgdYYY, attempting to rewind a specific deployment sgdXXX will result with:

• sgdXXX is paused
• sgdYYY is rewound
• sgdXXX is unpaused

This will corrupt sgdYYY because there is a graph-node instance that will keep indexing the subgraph while the rewind is in progress.

Relevant log output

graphman -c /etc/graph-node/config.toml info QmbeQwnRSX4Fs9Bo3LZCAvkb4psdmcS8AqKv78Tcq6ohED
Namespace | sgd4431 [shard19]
Hash      | QmbeQwnRSX4Fs9Bo3LZCAvkb4psdmcS8AqKv78Tcq6ohED
Version   | indexer-agent/78Tcq6ohED (current)
Chain     | arbitrum-one
Node ID   | index_node_2
Active    | true
----------+----------------------------------------------------------------------
Namespace | sgd4472 [shard20]
Hash      | QmbeQwnRSX4Fs9Bo3LZCAvkb4psdmcS8AqKv78Tcq6ohED
Version   | indexer-agent/78Tcq6ohED (current)
Chain     | arbitrum-one
Node ID   | index_node_4
Active    | false

graphman -c /etc/graph-node/config.toml rewind sgd4472  --block-number 363536000 --block-hash 0xca55ccd3f015993e059d786ecca9f8f02d64e43e3f32e85029fdc9b4d64896bd
Checking if its safe to rewind deployments
Pausing deployments
deployment QmbeQwnRSX4Fs9Bo3LZCAvkb4psdmcS8AqKv78Tcq6ohED[4472] is already paused

Waiting 20s to make sure pausing was processed

Rewinding deployments
  ... rewound QmbeQwnRSX4Fs9Bo3LZCAvkb4psdmcS8AqKv78Tcq6ohED[4472]
Resuming deployments
resuming QmbeQwnRSX4Fs9Bo3LZCAvkb4psdmcS8AqKv78Tcq6ohED[4472]
Operation completed

IPFS hash

No response

Subgraph name or link to explorer

No response

Some information to help us out

• Tick this box if this bug is caused by a regression found in the latest release.
• Tick this box if this bug is specific to the hosted service.
• I have searched the issue tracker to make sure this issue is not a duplicate.

OS information

None