<html><head></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; ">Hi Philip,<div><br></div><div>I do have about the same problem that you describe. There is my setup:</div><div><br></div><div>Gluster: Two bricks running gluster 3.2.6</div><div><br></div><div>Clients:</div><div>4 clients running native gluster fuse client.</div><div>2 clients running nfs client</div><div><br></div><div>My nfs client are not doing that much traffic but i was able to view after a couple days that the brick used to mount the nfs is having memory issue.</div><div><br></div><div>i can provide more info as needed to help correct the problem.</div><div><br></div><div>Thank's</div><div><br></div><div>Xavier</div><div><br></div><div><img height="427" width="501" apple-width="yes" apple-height="yes" id="78a75a38-371b-4ee6-a9b6-70accdb60cc6" src="cid:38664A7C-732E-4B64-AFF3-13FFDB09C17B@ldm.local"></div><div><br></div><div><br></div><div><div><div>Le 2012-06-12 à 08:18, Philip Poten a écrit :</div><br class="Apple-interchange-newline"><blockquote type="cite"><div>2012/6/12 Dan Bretherton <<a href="mailto:d.a.bretherton@reading.ac.uk">d.a.bretherton@reading.ac.uk</a>>:<br><blockquote type="cite">I wonder if this memory leak is the cause of the NFS performance degradation<br></blockquote><blockquote type="cite">I reported in April.<br></blockquote><br>That's probable, since the performance does go down for us too when<br>the glusterfs process reaches a large percentage of RAM. My initial<br>guess was that it's the file system cache that's being eradicated,<br>thus iowait increases. But a closer look at our munin graphs implies,<br>that it's also the user space that eats more and more CPU<br>proportionally with RAM:<br><br><a href="http://imgur.com/a/8YfhQ">http://imgur.com/a/8YfhQ</a><br><br>There are two restarts of the whole gluster process family visible on<br>those graphs: one a week ago at the very beginning (white in the<br>memory graph, as munin couldn't fork all it needed), and one<br>yesterday. The drop between 8 and 9 was due to a problemv unrelated to<br>gluster.<br><br>Pranith: I just made one dump, tomorrow I'll make one more and mail<br>them both to you so that you can compare them. While I just restarted<br>yesterday, the leak should be visible, as the process grows a few<br>hundred MB every day.<br><br>thanks for the fast reply,<br>Philip<br>_______________________________________________<br>Gluster-users mailing list<br>Gluster-users@gluster.org<br>http://gluster.org/cgi-bin/mailman/listinfo/gluster-users<br></div></blockquote></div><br></div></body></html>