| From: | Neil Conway <neilc(at)samurai(dot)com> |
|---|---|
| To: | Gregory Stark <stark(at)enterprisedb(dot)com> |
| Cc: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, Umar Farooq Minhas <umarfm13(at)hotmail(dot)com>, pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: Estimating seq_page_fetch and random_page_fetch |
| Date: | 2007-03-08 22:39:38 |
| Message-ID: | 1173393578.6422.7.camel@localhost.localdomain |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Thu, 2007-03-08 at 17:35 +0000, Gregory Stark wrote:
> When I was running tests I did it on a filesystem where nothing else was
> running. Between tests I unmounted and remounted it. As I understand it Linux
> associates the cache with the filesystem and not the block device and discards
> all pages from cache when the filesystem is unmounted.
On recent Linux kernels, /proc/sys/vm/drop_caches can also be useful:
http://linux.inet.hr/proc_sys_vm_drop_caches.html
You could also use posix_fadvise() to achieve a similar effect on a per-file
basis.
-Neil
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Gavin Sherry | 2007-03-08 23:33:48 | Re: Stream bitmaps |
| Previous Message | Shane Ambler | 2007-03-08 22:01:10 | Re: Auto creation of Partitions |