<br><div class="gmail_quote">On Sat, Jun 18, 2011 at 6:31 PM, <span dir="ltr"><<a href="mailto:gluster1206@akxnet.de">gluster1206@akxnet.de</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
Hi!<br>
<br>
> After upgrading to 3.2.1. and turning quick-read off<br>
> (<a href="http://bugs.gluster.com/show_bug.cgi?id=3011" target="_blank">http://bugs.gluster.com/show_bug.cgi?id=3011</a>) glusterfs works very fine<br>
> at the moment :-)<br>
<br>
After applying the proposed changes, Gluster seemed to work quite well<br>
during the last day.<br>
<br>
For safety reasons, the gluster partition was yet not used in<br>
production, the desired data on it was kept up to date with rsync.<br>
<br>
Taking it into production on a backup server, I noticed sudden load rises.<br>
<br>
Examining the problem, I discovered that a df freezes sometimes when<br>
reaching the gluster volumes and pauses some minutes and more until the<br>
information is shown. The df task does not show up in "top".<br><br></blockquote><div><br></div><div>How long does df pause for you? Also, can you post what's in the client logs?</div><div><br></div><div>Avati</div>
</div>