Re: Increase checkpoint segments ?

From: "Tomeh, Husam" <htomeh(at)firstam(dot)com>
To: "Abu Mushayeed" <abumushayeed(at)yahoo(dot)com>, pgsql-admin(at)postgresql(dot)org
Subject: Re: Increase checkpoint segments ?
Date: 2006-12-12 18:24:41
Message-ID: F1B0F9305B343E43A1C3EECE48B853D5093114@CITGSNA01SXCH02.ana.firstamdata.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-admin

Typically in this scenario, more and larger WAL segments can help speed
up your updates. However, the default size of a WAL segment is 16MB
which can be set at postgresql compile time, and can not be changed
thereafter. So, you'd need to increase the checkpoint_segments
parameters to speed up your updates. Increasing the WAL buffers which
are memory pages allocated in the shared memory for WAL data won't help.


Sincerely,

--
Husam
________________________________

From: pgsql-admin-owner(at)postgresql(dot)org
[mailto:pgsql-admin-owner(at)postgresql(dot)org] On Behalf Of Abu Mushayeed
Sent: Tuesday, December 12, 2006 9:17 AM
To: pgsql-admin(at)postgresql(dot)org
Subject: [ADMIN] Increase checkpoint segments ?

Hello,

I have many updates happening in the DB. All of them works with millions
of rows. When this happens I get the following message:

checkpoints are occurring too frequently (294 seconds apart)
Consider increasing the configuration parameter "checkpoint_segments".

Some values from my postgresql.conf files are as follows. My question is
do I really increase the checkpoint segments or should I increase the
WAL buffers or do an create table as and then insert instead of update
statement?

shared_buffers = 20000 #60000 # min 16 or
max_connections*2, 8KB each -- Bizgres work_mem = 65536 #131072 #65536
# min 64, size in KB -- Bizgres Database change from 1024 to 65536
maintenance_work_mem = 524288 #131072 # min 1024, size in KB
max_fsm_pages = 8000000 # min max_fsm_relations*16, 6
bytes each
max_fsm_relations = 32768 # min 100, ~70 bytes each

fsync = on # turns forced synchronization
on or off
wal_buffers = 128 # min 4, 8KB each
checkpoint_segments = 256 # in logfile segments, min 1,
16MB each
checkpoint_timeout = 3600 #300 # range 30-3600, in seconds
checkpoint_warning = 300 # in seconds, 0 is off
default_statistics_target = 250 # range 1-1000

Thanks
Abu

________________________________

Have a burning question? Go to Yahoo! Answers
<http://answers.yahoo.com/;_ylc=X3oDMTFvbGNhMGE3BF9TAzM5NjU0NTEwOARfcwMz
OTY1NDUxMDMEc2VjA21haWxfdGFnbGluZQRzbGsDbWFpbF90YWcx> and get answers
from real people who know.

**********************************************************************
This message contains confidential information intended only for the use of the addressee(s) named above and may contain information that is legally privileged. If you are not the addressee, or the person responsible for delivering it to the addressee, you are hereby notified that reading, disseminating, distributing or copying this message is strictly prohibited. If you have received this message by mistake, please immediately notify us by replying to the message and delete the original message immediately thereafter.

Thank you.

FADLD Tag
**********************************************************************

In response to

Browse pgsql-admin by date

  From Date Subject
Next Message Scott Marlowe 2006-12-12 18:38:33 Re: Hide structure
Previous Message Rajesh Kumar Mallah 2006-12-12 17:57:25 Re: query crashes 8.2.0 but not 8.1.5