Re: Fatal: 2149478 (UNCLASSIFIED) (UNCLASSIFIED)

Classification: UNCLASSIFIED
Caveats: NONE

I am currently working on figuring out the size of my systems. It looks to
me like I am dealing with 2 gig systems.. Although on our road servers (we
travel) we are dealing with 12 gig systems. Unfortunately I have had hit
or miss problems with these work_mem and maintenance_mem calculations.. On
some systems they work.. On some they don't. Here is the calculations that
we are using (these were developed by a employee that no longer works here).

Shared_buffers = (total_mem_mb * 1024 * .06) / 8
SHMMAX = (shared_buffers * 8 * 1.2) * 1024
SHMALL = (total_mem_mb * 1024 *4)
WORK_MEM = (free_mem_mb * 1024 *.02)
MAINTENANCE_WORK_MEM = (free_mem_mb * 1024 * .3)

-----Original Message-----
From: Tom Lane [mailto:tgl(at)sss(dot)pgh(dot)pa(dot)us]
Sent: Thursday, May 11, 2006 12:29 PM
To: Joshua D. Drake
Cc: Crowson, Sarah J Ms (Contractor) Northrop Grumman;
'pgsql-admin(at)postgresql(dot)org'
Subject: Re: [ADMIN] Fatal: 2149478 (UNCLASSIFIED)

"Joshua D. Drake" <jd(at)commandprompt(dot)com> writes:
>> Fatal: 2149478 is outside the valid range for parameter
>> "maintenance_work_mem" ............

> That is odd. I just testing on 8.1.3:

> template1=# set maintenance_work_mem = 2149478; SET

I'm betting Joshua is testing on a 64-bit machine. Since 8.1, the max
allowed value is dependent on sizeof(pointer) ...

regards, tom lane
Classification: UNCLASSIFIED
Caveats: NONE