Alvaro Herrera <alvherre(at)commandprompt(dot)com> writes:
> Tom Lane wrote:
>> I couldn't reproduce it in a few tries. A reasonable guess is that
>> it's triggered by autovacuum deciding to vacuum the table sometime
>> before the VACUUM FULL starts. Anyone want to try to reproduce it?
> Hum, aren't vacuums supposed to be blocked by each other?
Sure. I'm not thinking it's a case of concurrent vacuums (if it is,
we've got worse problems than anyone imagined), but rather that the
autovac left the table in a state that exposes a bug in the subsequent
VACUUM FULL. Since we've whacked the tqual.c logic around recently,
the problem might actually lie there...
regards, tom lane