1986 - Fix duplicate pool stats

[Cmdv]

Description

This fixes #1986

Due to complexity of partial entries and rollbacks it was just as performant to add a unique constraints to the table

Checklist

• Commit sequence broadly makes sense
• Commits have useful messages
• New tests are added if needed and existing tests are updated
• Any changes are noted in the changelog
• Code is formatted with fourmolu on version 0.10.1.0 (which can be run with scripts/fourmolize.sh)
• Self-reviewed the diff

Migrations

• The pr causes a breaking change of type a,b or c
• If there is a breaking change, the pr includes a database migration and/or a fix process for old values, so that upgrade is possible
• Resyncing and running the migrations provided will result in the same database semantically

If there is a breaking change, especially a big one, please add a justification here. Please elaborate
more what the migration achieves, what it cannot achieve or why a migration is not possible.

[kderme]

This adds a unique key, but doesn't change the function that inserts the values. So if dbsync tries again to insert a duplicate, it will crash, instead of doing a no-op.

Also, instances that upgrade to this version and already have duplicates, would fail to create the unique key altoether.

Finally, any idea what could be the root cause of creating a duplicate in the first place?

[Cmdv]

ok think I've found it:

onlyDelete "PoolStat" [PoolStatEpochNo >. epochN] should be onlyDelete "PoolStat" [PoolStatEpochNo >=. epochN]

cardano-db-sync/cardano-db/src/Cardano/Db/Operations/Delete.hs

Lines 122 to 138 in e00405d

	deleteUsingEpochNo :: MonadIO m => Word64 -> ReaderT SqlBackend m [(Text, Int64)]
	deleteUsingEpochNo epochN = do
	countLogs <-
	concat
	<$> sequence
	[ onlyDelete "Epoch" [EpochNo ==. epochN]
	, onlyDelete "DrepDistr" [DrepDistrEpochNo >. epochN]
	, onlyDelete "RewardRest" [RewardRestSpendableEpoch >. epochN]
	, onlyDelete "PoolStat" [PoolStatEpochNo >. epochN]
	]
	nullLogs <- do
	a <- setNullEnacted epochN
	b <- setNullRatified epochN
	c <- setNullDropped epochN
	e <- setNullExpired epochN
	pure [("GovActionProposal Nulled", a + b + c + e)]
	pure $ countLogs <> nullLogs

Will write tests see if that's the case

[Cmdv]

@kderme should we do a check to see if there are duplicates when starting DB sync and if so delete them? This should cover those who have old corrupted data