Re: Interesting VACUUM notice

Mario Weilguni <mweilguni(at)sime(dot)com> writes:
> I found a message I've never seen before in VACUUM, its:
> NOTICE: Too old parent tuple found - can't continue repair_frag

> This sounds like a solution should be available, but it seems to
> happen anyway. Yesterday I've found no way to fix this problem, but
> today it's not reproduceable any more. Might this notice indicate a
> serious problem?

No; it's just an unimplemented case in VACUUM FULL's tuple-chain-moving
code. The problem disappears when the chain stops being interesting --
that is, as soon as existing transactions finish so that no transaction
can see the older states of that row anymore. Meanwhile, the only
consequence is that VACUUM may not be able to shrink the table as much
as it otherwise could've.

I downgraded all those "can't continue repair_frag" notices to DEBUG1
level yesterday, since they are expected (though improbable) cases.

regards, tom lane