On Wed, Sep 27, 2023 at 10:29:25AM -0700, Andres Freund wrote:
> I don't think going for size_t is a viable path for fixing this. I'm pretty
> sure the initial patch would trigger a type mismatch from guc_tables.c - we
> don't have infrastructure for size_t GUCs.
Nothing marked as PGDLLIMPORT uses size_t in the tree currently, FWIW.
> Perhaps we ought to error out (in BackendStatusShmemSize() or such) if
> pgstat_track_activity_query_size * MaxBackends >= 4GB?
Yeah, agreed that putting a check like that could catch errors more
quickly.
> Frankly, it seems like a quite bad idea to have such a high limit for
> pgstat_track_activity_query_size. The overhead such a high value has will
> surprise people...
Still it could have some value for some users with large analytical
queries where the syslogger is not going to be a bottleneck? It seems
too late to me to change that, but perhaps the docs could be improved
to tell that using a too high value can have performance consequences,
while mentioning the maximum value.
--
Michael