Bruce Momjian <bruce@momjian.us> writes:
> Tom Lane wrote:
>> Not really, but maybe it would be sensible to reset last_pgstat_start_time
>> when doing a database-wide restart?
> You mean like this, attached?
I'd be fairly surprised if resetting the variable in the child process
had any effect on the postmaster...
I think a correct fix would involve exposing either the variable or a
routine to zero it from pgstat.c, and having postmaster.c do that at
the points where it intentionally SIGQUITs the stats collector.
regards, tom lane