On Thu, Jan 9, 2014 at 6:36 AM, Gabriele Bartolini
<gabriele.bartolini@2ndquadrant.it> wrote:
> Il 08/01/14 18:42, Simon Riggs ha scritto:
>> Not sure I see why it needs to be an SRF. It only returns one row.
> Attached is version 4, which removes management of SRF stages.
I have been looking at v4 a bit more, and found a couple of small things:
- a warning in pgstatfuncs.c
- some typos and format fixing in the sgml docs
- some corrections in a couple of comments
- Fixed an error message related to pg_stat_reset_shared referring
only to bgwriter and not the new option archiver. Here is how the new
message looks:
=# select pg_stat_reset_shared('popo');
ERROR: 22023: unrecognized reset target: "popo"
HINT: Target must be "bgwriter" or "archiver"
A new version is attached containing those fixes. I played also with
the patch and pgbench, simulating some archive failures and successes
while running pgbench and the reports given by pg_stat_archiver were
correct, so I am marking this patch as "Ready for committer".
Regards,
--
Michael