xfs
[Top] [All Lists]

Re: [PATCH 09/10] xfs: xfs_fs_write_inode() can fail to write inodes syn

To: bpm@xxxxxxx
Subject: Re: [PATCH 09/10] xfs: xfs_fs_write_inode() can fail to write inodes synchronously V2
From: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Date: Wed, 3 Feb 2010 15:55:33 -0500
Cc: Christoph Hellwig <hch@xxxxxxxxxxxxx>, Dave Chinner <david@xxxxxxxxxxxxx>, xfs@xxxxxxxxxxx
In-reply-to: <20100203180706.GH5702@xxxxxxx>
References: <1265153104-29680-1-git-send-email-david@xxxxxxxxxxxxx> <1265153104-29680-10-git-send-email-david@xxxxxxxxxxxxx> <20100203112753.GA19996@xxxxxxxxxxxxx> <20100203180706.GH5702@xxxxxxx>
User-agent: Mutt/1.5.19 (2009-01-05)
On Wed, Feb 03, 2010 at 12:07:06PM -0600, bpm@xxxxxxx wrote:
> Hey Christoph,
> 
> On Wed, Feb 03, 2010 at 06:27:53AM -0500, Christoph Hellwig wrote:
> > That leaves open the NFSD case thought.  I'd prefer to have that fixed
> > if possibly.  Ben, any chance you could send your patch to use fsync
> > to the nfs list ASAP?
> 
> I'll make some time to work on it.
> 
> > I think we'd be even better off to just force
> > -o wsync and disable ->write_inode entirely for NFS, any chance you
> > could test such a patch on your setup?
> 
> Sure.  IIRC, previous tests were w/ -o wsync and write_inode switched to
> fsync passing 1 for fdatasync.  I'll try this too.

-o wsync should be very similar to using ->fsync or Dave's new
->write_inode.  What the synchronous transaction does is causing a log
force, which is exactly what fsync does in case the inode is still
pinned (except not quite as optimal, but I'll send a patch for that)

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