From: | Heikki Linnakangas <hlinnakangas(at)vmware(dot)com> |
---|---|
To: | Christian Kruse <christian(at)2ndQuadrant(dot)com> |
Cc: | pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: Patch: show xid and xmin in pg_stat_activity and pg_stat_replication |
Date: | 2014-01-13 18:06:56 |
Message-ID: | 52D42B40.80904@vmware.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On 12/17/2013 04:58 PM, Christian Kruse wrote:
> attached you will find a patch for showing the current transaction id
> (xid) and the xmin of a backend in pg_stat_activty and the xmin in
> pg_stat_replication.
Docs.
When an admin is looking for a long-running transaction that's blocking
vacuum, he will currently rely on the timestamp fields, xact_start and
query_start. I'm not sure how much extra value this adds over those
timestamps in pg_stat_activity, but there are not such fields in
pg_stat_replication, so that part is definitely useful. And if we're
going to add xmin to pg_stat_replication, it makes sense to add it to
pg_stat_activity too. Unless someone can come up with something better
to display for walsenders. The timestamp of the last commit record
that's been replayed, perhaps?
What else would a user would want to do with these?
This definitely sounds useful to me as a developer, though. So I'm
thinking we should add these for that reason, in any case.
- Heikki
From | Date | Subject | |
---|---|---|---|
Next Message | Hannu Krosing | 2014-01-13 18:12:55 | Re: Standalone synchronous master |
Previous Message | Alexander Korotkov | 2014-01-13 18:01:20 | Re: PoC: Partial sort |