> 27 июня 2019 г., в 16:38, Heikki Linnakangas <hlinnaka@iki.fi> написал(а):
>
> I haven't done any testing on this yet. Andrey, would you happen to have an environment ready to test this?
Patches do not deadlock and delete pages on "rescan test" - setup that we used to detect "left jumps" in during
developmentof physical vacuum. check-world is happy on my machine.
I really like that now there is GISTDeletedPageContents and we do not just cast *(FullTransactionId *)
PageGetContents(page).
But I have stupid question again, about this code:
https://github.com/x4m/postgres_g/commit/096d5586537d29ff316ca8ce074bbe1b325879ee#diff-754126824470cb8e68fd5e32af6d3bcaR417
nextFullXid = ReadNextFullTransactionId();
diff = U64FromFullTransactionId(nextFullXid) -
U64FromFullTransactionId(latestRemovedFullXid);
if (diff < MaxTransactionId / 2)
{
TransactionId latestRemovedXid;
// sleep(100500 hours); latestRemovedXid becomes xid from future
latestRemovedXid = XidFromFullTransactionId(latestRemovedFullXid);
ResolveRecoveryConflictWithSnapshot(latestRemovedXid,
xlrec->node);
}
Do we have a race condition here? Can latestRemovedXid overlap and start to be xid from future?
I understand that it is purely hypothetical, but still latestRemovedXid is from ancient past already.
Best regards, Andrey Borodin.