| From: | Alvaro Herrera <alvherre(at)2ndquadrant(dot)com> |
|---|---|
| To: | Bryce Nesbitt <bryce2(at)obviously(dot)com> |
| Cc: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, pgsql-bugs(at)postgresql(dot)org |
| Subject: | Re: BUG #12834: Avoiding Disk Control Writes for better Laptop / SSD compatibility. pgstats |
| Date: | 2015-03-05 19:44:57 |
| Message-ID: | 20150305194457.GU3291@alvh.no-ip.org |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-bugs |
Bryce Nesbitt wrote:
> Is autovac writing anything of import during its periodic wakeup on an
> idle cluster?
Autovacuum is not writing anything. What it is doing is asking whether
anyone has written anything -- and the stats collector responds by
writing these stat files.
The issue is that the timestamp-of-last-update bit, which is critical to
autovacuum, does not get written unless the whole file is. I guess it
is possible to optimize that, so that the file is not touched other than
the timestamp if the database has been completely idle. I have no idea
how complex this is, or whether this scenario is really worth
optimizing.
--
Álvaro Herrera http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Ha Quan Le | 2015-03-05 22:50:05 | Re: BUG #12821: Cannot install xlogdump |
| Previous Message | Bryce Nesbitt | 2015-03-05 19:37:58 | Re: BUG #12834: Avoiding Disk Control Writes for better Laptop / SSD compatibility. pgstats |