| From: | Bruce Momjian <bruce(at)momjian(dot)us> |
|---|---|
| To: | Greg Stark <stark(at)mit(dot)edu> |
| Cc: | Merlin Moncure <mmoncure(at)gmail(dot)com>, Andres Freund <andres(at)anarazel(dot)de>, Jeff Janes <jeff(dot)janes(at)gmail(dot)com>, "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org>, Julien Rouhaud <julien(dot)rouhaud(at)dalibo(dot)com> |
| Subject: | Re: Allow a per-tablespace effective_io_concurrency setting |
| Date: | 2015-09-04 16:23:38 |
| Message-ID: | 20150904162338.GB26916@momjian.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers pgsql-performance |
On Fri, Sep 4, 2015 at 05:21:38PM +0100, Greg Stark wrote:
> Wouldn't SSDs need much *less* aggressive prefetching? There's still
> latency and there are multiple I/O channels so they will still need
> some. But spinning media gives latencies measured in milliseconds. You
> can process a lot of tuples in milliseconds. If you have a hundred
> spindles you want them all busy doing seeks because in the 5ms it
> takes them to do that you can proess all the results on a single cpu
> and the rest of time is spend waiting.
>
> When your media has latency on the order of microseconds then you only
> need to have a small handful of I/O requests in flight to keep your
> processor busy.
Well, there is still the processing time of getting that data ready.
All I know is that people have reported that prefetching is even more
useful for SSDs.
--
Bruce Momjian <bruce(at)momjian(dot)us> http://momjian.us
EnterpriseDB http://enterprisedb.com
+ Everyone has their own god. +
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andres Freund | 2015-09-04 16:37:25 | Re: Allow a per-tablespace effective_io_concurrency setting |
| Previous Message | Bruce Momjian | 2015-09-04 16:22:06 | Re: Freeze avoidance of very large table. |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andres Freund | 2015-09-04 16:37:25 | Re: Allow a per-tablespace effective_io_concurrency setting |
| Previous Message | Greg Stark | 2015-09-04 16:21:38 | Re: Allow a per-tablespace effective_io_concurrency setting |