[Date Prev][Date Next] [Thread Prev][Thread Next]
[Thread Index]
[Date Index]
[Author Index]
Re: [Maybe Spam] Re: memory cache problem
- From: Al Tobey <albert tobey priority-health com>
- To: "Discussion of Red Hat Enterprise Linux 3 (Taroon)" <taroon-list redhat com>
- Subject: Re: [Maybe Spam] Re: memory cache problem
- Date: Wed, 15 Sep 2004 13:26:38 -0400
On Wed, 2004-09-15 at 11:59, John Haxby wrote:
> Mark Dadgar wrote:
>
> > This really, really sucks. And hopefully will get fixed one of these
> > days, as it's going to be a real problem soon when I go into
> > production. Paging out the SQL server is a very bad idea.
I think recent versions of MySQL can be instructed to mlock() it's
memory so it can't be swapped. PostgreSQL probably has this as well.
Oracle, I think, does this by default.
>
> Ahh. I see. I do remember something in the archives about tuning the
> cache parameters for precisely this reason -- another poster (sorry, the
> message has been deleted) suggested doing exactly the same thing. I
> agree that you don't want the SQL server paging out even if it appears
> to be idle. I wish I could remember something else about the tuning,
> but there was definitely something about SQL servers and cache tuning on
> the list a little while back.
> jch
Grab the Documentation/filesystems/proc.txt file from a 2.6 kernel
distribution - the one in the RH kernel source is severely outdated.
Mainly, you'll want to tweak vfs_cache_pressure (vm.vfs_cache_pressure
in /etc/sysctl.conf) on RHEL. It appears that some of the tunables for
percentage of memory to use for caching are no longer available in RHEL
and 2.6 kernels.
vfs_cache_pressure
------------------
Controls the tendency of the kernel to reclaim the memory which is used for
caching of directory and inode objects.
At the default value of vfs_cache_pressure=100 the kernel will attempt to
reclaim dentries and inodes at a "fair" rate with respect to pagecache and
swapcache reclaim. Decreasing vfs_cache_pressure causes the kernel to prefer
to retain dentry and inode caches. Increasing vfs_cache_pressure beyond 100
causes the kernel to prefer to reclaim dentries and inodes.
** ** ** PRIVILEGED AND CONFIDENTIAL ** ** **
This email transmission contains privileged and confidential information
intended only for the use of the individual or entity named above. Any
unauthorized review, use, disclosure or distribution is prohibited and
may be a violation of law. If you are not the intended recipient or a
person responsible for delivering this message to an intended recipient,
please delete the email and immediately notify the sender via the email
return address or mailto:postmaster priority-health com Thank you.
[Date Prev][Date Next] [Thread Prev][Thread Next]
[Thread Index]
[Date Index]
[Author Index]