xfs
[Top] [All Lists]

Re: Exponential memory usage?

To: Tap <tap@xxxxxxxxxxxxxxxx>, xfs@xxxxxxxxxxx
Subject: Re: Exponential memory usage?
From: Eric Sandeen <sandeen@xxxxxxxxxxx>
Date: Fri, 28 Mar 2014 20:24:24 -0500
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <DF4B0D0E-6D5A-4670-9301-A1DCD0EEAE6E@xxxxxxxxxxxxxxxx>
References: <DF4B0D0E-6D5A-4670-9301-A1DCD0EEAE6E@xxxxxxxxxxxxxxxx>
User-agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.9; rv:24.0) Gecko/20100101 Thunderbird/24.4.0
On 3/28/14, 12:53 PM, Tap wrote:
> I have a Linux CentOs-based (6.5) system, running Linux 3.10.29
> beneath xen 4.2.3-26 with a raid array as follows:

<snip>

> Anyway at some point the system had become unusable.  Something as
> simple as:
> 
> find  /raid -type f -- or -- ls -lR /raid
> 
> Would walk the entire system out of RAM.  Looking at slabtop it looks
> like this is due mostly to xfs_inode memory usage.  Note that since
> these problems began I stopped running all sub-ordinate domains and
> am now only running dom0.  In fact I've allocated all 32 GB to that
> domain, and memory problems still persist.

Out of?  Did the OOM-killer actually show up, or did you simply see
memory usage rise?

The dcache is tenacious.  dentries are cached pretty aggressively, and
those in turn will pin the vfs/xfs inodes.

But memory is there to be used; unless you actually see problems,
then the system is likely working as it should be.

<snip>

> Note:  FWIW Un-mounting the filesystem also recovers the memory.

Yep, that would clear out the cache.

-Eric

<Prev in Thread] Current Thread [Next in Thread>