On 22/03/2014, at 8:48 AM, Emmanuel Dreyfus wrote: > Vijay Bellur <vbel...@redhat.com> wrote: >> Can you please try with the latest release-3.5 code? In my testing >> (untar + rm -rf ) does not cause memory leaks after [1] was merged. > > It completed my NetBSD build test (see below) on 2x2 > replicated/distributed setup. I did not tried rebooting bricks randomly > during the test, this is my second test, but I guess there is something > more important to focus on right now: > > The bricks glusterfsd processes after the test use respectively 1.29 GB, > 2.14 GB, 2.81 GB, and 1.49 GB of virtual memory. This means there is > still a huge memory leak left. > > On the client side, the glusterfsd uses 911 MB, which looks huge too.
Out of curiosity, how much memory and how many cpu's do the servers have? Asking because although Rackspace doesn't natively support NetBSD, it might be possible to sleaze :) an install there anyway, so we can replicate the problem and try to identify it. If it works, might as well try to match your cpu's and ram too. :) Regards and best wishes, Justin Clift -- Open Source and Standards @ Red Hat twitter.com/realjustinclift _______________________________________________ Gluster-devel mailing list Gluster-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/gluster-devel
