| From: | "Luke Lonergan" <llonergan(at)greenplum(dot)com> |
|---|---|
| To: | "Merlin Moncure" <mmoncure(at)gmail(dot)com>, "Worky Workerson" <worky(dot)workerson(at)gmail(dot)com> |
| Cc: | "Markus Schaber" <schabi(at)logix-tt(dot)com>, pgsql-performance(at)postgresql(dot)org |
| Subject: | Re: Best COPY Performance |
| Date: | 2006-10-25 16:34:01 |
| Message-ID: | C164E009.51ED%llonergan@greenplum.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
Merlin,
On 10/25/06 8:38 AM, "Merlin Moncure" <mmoncure(at)gmail(dot)com> wrote:
> in theory, with 10 10k disks in raid 10, you should be able to keep
> your 2fc link saturated all the time unless your i/o is extremely
> random. random i/o is the wild card here, ideally you should see at
> least 2000 seeks in bonnie...lets see what comes up.
The 2000 seeks/sec are irrelevant to Postgres with one user doing COPY.
Because the I/O is single threaded, you will get one disk worth of seeks for
one user, roughly 150/second on a 10K RPM drive.
I suspect the problem here is the sequential I/O rate - let's wait and see
what the dd test results look like.
- Luke
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Craig A. James | 2006-10-25 16:52:14 | Re: Best COPY Performance |
| Previous Message | Carlo Stonebanks | 2006-10-25 15:46:41 | commit so slow program looks frozen |