"Merlin Moncure" <merlin(dot)moncure(at)rcsonline(dot)com> writes:
>> I was also wondering about adding a current-XID column to
>> pg_stat_activity, and encouraging people to join pg_locks and
>> pg_stat_activity on XID instead of PID.
> That would be awesome. Is there any performance penalty to do this?
I gave up on the idea after I realized that current XID tends to change
a lot faster than the pg_stats mechanism is designed to track.
(Consider the half-second lag for starters...) Turning pg_stats into a
realtime mechanism would be horridly expensive.
regards, tom lane