xfs
[Top] [All Lists]

Re: Consistent throughput challenge -- fragmentation?

To: stan@xxxxxxxxxxxxxxxxx
Subject: Re: Consistent throughput challenge -- fragmentation?
From: Brian Cain <brian.cain@xxxxxxxxx>
Date: Mon, 25 Feb 2013 16:06:55 -0600
Cc: xfs@xxxxxxxxxxx
Delivered-to: xfs@xxxxxxxxxxx
Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:x-received:in-reply-to:references:date:message-id :subject:from:to:cc:content-type; bh=sYQSJ6DgA/R+CBMdohQXXmdUTl8Mv3w/zp6kQjPgFBU=; b=n7dhvUwbQjiBGOj9j6Ih+x5zIH+d5VOVCci7iqxUWZ1puRdfFc86T3Zd0kkyDSqjFB NZwjFfB6H1QZJjgbv5Yum9V0zO1EtFPh5lx7/GL21gtOSLhY2NQVFSZVSRm/QsV2P7ej lBfADHJ79J4qJctdYCxFzEZpDSvkRVwltIALTxfKaG7w3O8n20OHxUdhps3hgkE/16/F guiwGxtgbLda7E/yfUBwbPBjnOKkl6qwxXnG33OW0KGy9lbhqodzyqO5sS0X4K4i2St3 4xDXWNI6VCTcx+MIcveCLEMlaOYQMfQZVqT+6VomTUZwUKjknvQOk+2yzpFr5djoGQSp zR0Q==
In-reply-to: <512BDA2A.5050600@xxxxxxxxxxxxxxxxx>
References: <CAEWpfG_DKJt1MmWS1tARH4OmYwpSt=A-DzwKkGcD67LuR6k=Bg@xxxxxxxxxxxxxx> <512BDA2A.5050600@xxxxxxxxxxxxxxxxx>



On Mon, Feb 25, 2013 at 3:39 PM, Stan Hoeppner <stan@xxxxxxxxxxxxxxxxx> wrote:
>
>
> > Can anyone provide any suggestions as to an explanation for the behavior or
> > a way to mitigate it?  Running xfs_fsr didn't seem to improve the results.
>
> The usual cause of such aged filesystem low performance is free space
> fragmentation.  xfs_fsr will defragment files, but in doing so it
> *increases* free space fragmentation, thus won't help the situation.
>
> > I'm happy to share benchmarks, specific results data, or describe the
> > hardware being used for the measurements if it's helpful.
>
> Paste the output of 'xfs_db -r -c freesp /dev/[device]' just before you
> do the large file write.  This will show us the free space distribution
> histogram.
>

Running now...

Here's a single sample:

   from      to extents  blocks    pct
      1       1     128     128   0.00
      2       3       6      18   0.00
      4       7       1       7   0.00
      8      15      30     275   0.00
    512    1023       1     528   0.01
   2048    4095       1    2656   0.03
4194304 8388608       1 8388588  99.96

Not sure whether the cycle following this output experienced "only good" results or if it included poor performing samples too.  Is it only useful to see the "freesp" output in cases where the poor performance occurred?

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