Peter Eisentraut <peter_e@gmx.net> writes:
> On lör, 2012-05-12 at 12:59 -0400, Tom Lane wrote:
>> Now it's entirely likely that there is nobody out there relying on
>> such a thing, but nonetheless this is a compatibility break, and an
>> unnecessary one IMO. You haven't shown any convincing reason why we
>> need to change the behavior of age() on master servers at all.
> Recall that this thread originally arose out of age() being called by a
> monitoring tool. It would be nice if repeatedly calling age() on an
> otherwise idle database would not change the result. Currently, you
> would never get a "stable" state on such a check, and moreover, you
> would not only get different results but different long-term behavior
> between master and standby.
Hm. Interesting argument, but why exactly would you expect that age()
would work differently from, say, wall clock time? And how likely is it
that a database that requires monitoring is going to have exactly zero
transactions over a significant length of time?
(In any case, my primary beef at the moment is not with whether it's a
good idea to change age()'s behavior going forward, but rather with
having back-patched such a change.)
regards, tom lane