Jan Wieck <JanWieck@Yahoo.com> writes:
> I guess nobody ever relied that heavily on data to be persistent at the
> microsecond the NOTIFY arrives ...
Sure they have.
In theory you cannot see a NOTIFY before the sending transaction
commits, because the sender is holding a lock on pg_notify and you can't
even find out that you've been notified until he releases it.
Your idea that the COMMIT WAL record is getting dropped would fit the
facts, but I really am having a hard time believing it. Why would the
COMMIT record be more prone to loss than any other record? All the
cases I have seen so far have the right number of non-COMMIT records in
the log, so the bogus transaction is not getting lost altogether.
regards, tom lane