| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Heikki Linnakangas <heikki(at)enterprisedb(dot)com> |
| Cc: | PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org>, Greg Smith <gsmith(at)gregsmith(dot)com>, Hannu Krosing <hannu(at)skype(dot)net>, ITAGAKI Takahiro <itagaki(dot)takahiro(at)oss(dot)ntt(dot)co(dot)jp>, Greg Stark <greg(dot)stark(at)enterprisedb(dot)com> |
| Subject: | Re: Controlling Load Distributed Checkpoints |
| Date: | 2007-06-07 14:16:25 |
| Message-ID: | 719.1181225785@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers pgsql-patches |
Heikki Linnakangas <heikki(at)enterprisedb(dot)com> writes:
> Thinking about this whole idea a bit more, it occured to me that the
> current approach to write all, then fsync all is really a historical
> artifact of the fact that we used to use the system-wide sync call
> instead of fsyncs to flush the pages to disk. That might not be the best
> way to do things in the new load-distributed-checkpoint world.
> How about interleaving the writes with the fsyncs?
I don't think it's a historical artifact at all: it's a valid reflection
of the fact that we don't know enough about disk layout to do low-level
I/O scheduling. Issuing more fsyncs than necessary will do little
except guarantee a less-than-optimal scheduling of the writes.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | ohp | 2007-06-07 15:01:19 | little PITR annoyance |
| Previous Message | Heikki Linnakangas | 2007-06-07 12:23:06 | Re: Controlling Load Distributed Checkpoints |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Jim C. Nasby | 2007-06-07 15:50:36 | Re: Autovacuum launcher doesn't notice death of postmaster immediately |
| Previous Message | Heikki Linnakangas | 2007-06-07 12:23:06 | Re: Controlling Load Distributed Checkpoints |