Thread: about shared_buffer and kernel.shmmax

about shared_buffer and kernel.shmmax

From
wangqi
Date:
Hi everybody,
shared_buffer maximum limit is ?
I set shared_buffer more than 250G, I got error at startup pg.

------------------------------------------------------------------------
This error usually means that PostgreSQL's request for a shared memory
segment exceeded your kernel's SHMMAX parameter. You can either reduce
the request size or reconfigure the kernel with larger SHMMAX. To reduce
the request size (currently 439371702272 bytes), reduce PostgreSQL's
shared_buffers parameter (currently 52428800) and/or its max_connections
parameter (currently 104).
-------------------------------------------------------------------------

but now
kernel.shmmax = 966367641600
kernel.shmall = 235929600

 --
----------------------------------------------------
Yours sincerely,
Qi Wang
                         Edgesoft (China) Co.,Ltd
                         F26th Block B,Zhongli Building,
                         No.32 Huoju Road,
                         Hi-Tech Zone,Dalian,China
                         Zip:116023
                         Tel:+86-411-84753511
                         Fax:+86-411-84753577
                         http://www.edgesoft.cn
                         mailto:wangqi@edgesoft.cn
----------------------------------------------------


Re: about shared_buffer and kernel.shmmax

From
Craig Ringer
Date:
On 06/21/2012 11:38 AM, wangqi wrote:
Hi everybody,
shared_buffer maximum limit is ?
I set shared_buffer more than 250G, I got error at startup pg.



http://www.postgresql.org/docs/9.1/static/kernel-resources.html#SYSVIPC

http://www.postgresql.org/docs/9.1/static/runtime-config-resource.html

http://wiki.postgresql.org/wiki/Guide_to_reporting_problems

--
Craig Ringer

Re: about shared_buffer and kernel.shmmax

From
Greg Smith
Date:
On 06/20/2012 11:38 PM, wangqi wrote:
> shared_buffer maximum limit is ?
> I set shared_buffer more than 250G, I got error at startup pg.

There's not much evidence that values larger than 16GB are really
productive.  You really do not want put all their memory in just that
one place.

--
Greg Smith   2ndQuadrant US    greg@2ndQuadrant.com   Baltimore, MD
PostgreSQL Training, Services, and 24x7 Support www.2ndQuadrant.com