[Top] [All Lists]

Re: XFS blocked task in xlog_cil_force_lsn

To: Stefan Priebe - Profihost AG <s.priebe@xxxxxxxxxxxx>
Subject: Re: XFS blocked task in xlog_cil_force_lsn
From: Dave Chinner <david@xxxxxxxxxxxxx>
Date: Tue, 17 Apr 2012 18:32:10 +1000
Cc: stable@xxxxxxxxxxxxxxx, Greg KH <gregkh@xxxxxxxxxxxxxxxxxxx>, Christoph Hellwig <hch@xxxxxxxxxxxxx>, "xfs@xxxxxxxxxxx" <xfs@xxxxxxxxxxx>, Philipp Herz - Profihost AG <p.herz@xxxxxxxxxxxx>
In-reply-to: <4F8D27A1.2090809@xxxxxxxxxxxx>
References: <4F8BC0B9.90604@xxxxxxxxxxxx> <20120416235647.GA6734@dastard> <4F8D27A1.2090809@xxxxxxxxxxxx>
User-agent: Mutt/1.5.21 (2010-09-15)
On Tue, Apr 17, 2012 at 10:19:45AM +0200, Stefan Priebe - Profihost AG wrote:
> Am 17.04.2012 01:56, schrieb Dave Chinner:
> > Your first question should be: is this a known problem?
> > And the answer to that is: I don't think so.
> Sorry absolutely correct.
> > So, is the system actually hung?
> No. The kernel has ended the stalled request.
> > Were there IO errors around this time?
> No.
> > did the system recover and continue to work normally?
> Yes.
> > What's your storage hardware, application, workload, etc?
> It a Raid 10 of 1TB SATA Drives in front of a 3Ware Controller doing
> heavy rsync backups.

Hmmm. It sounds like it is just taking a long time for the IO to
complete, not that there is any other problem. If the IO is realy,
really slow, the a log force is definitely a place it coul dstall
for some time.

Does the RAID controller have a BBWC? What IO scheduler are you
using? Does mounting with nobarriers reduce the stall time? Can you
reduce the IO load so that it doesn't end up in this state?


Dave Chinner

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