From: | Haribabu kommi <haribabu(dot)kommi(at)huawei(dot)com> |
---|---|
To: | KONDO Mitsumasa <kondo(dot)mitsumasa(at)lab(dot)ntt(dot)co(dot)jp> |
Cc: | Amit Kapila <amit(dot)kapila16(at)gmail(dot)com>, Fujii Masao <masao(dot)fujii(at)gmail(dot)com>, PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: Compression of full-page-writes |
Date: | 2013-10-08 11:13:04 |
Message-ID: | 8977CB36860C5843884E0A18D8747B0372BC790D@szxeml558-mbs.china.huawei.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On 08 October 2013 15:22 KONDO Mitsumasa wrote:
> (2013/10/08 17:33), Haribabu kommi wrote:
>> The checkpoint_timeout and checkpoint_segments are increased to make sure no checkpoint happens during the test run.
>Your setting is easy occurred checkpoint in checkpoint_segments = 256. I don't know number of disks in your test server, in my test server which has 4 magnetic disk(1.5k rpm), postgres generates 50 - 100 WALs per minutes.
A manual checkpoint is executed before starting of the test and verified as no checkpoint happened during the run by increasing the "checkpoint_warning".
>And I cannot understand your setting which is sync_commit = off. This setting tend to cause cpu bottle-neck and data-loss. It is not general in database usage.
Therefore, your test is not fair comparison for Fujii's patch.
I chosen the sync_commit=off mode because it generates more tps, thus it increases the volume of WAL.
I will test with sync_commit=on mode and provide the test results.
Regards,
Hari babu.
From | Date | Subject | |
---|---|---|---|
Next Message | Rushabh Lathia | 2013-10-08 11:38:17 | Re: insert throw error when year field len > 4 for timestamptz datatype |
Previous Message | shailesh singh | 2013-10-08 10:55:59 | Urgent Help Required |