Red Hat Bugzilla – Bug 763983
NFS server hangs on 40 dds running in parallel
Last modified: 2011-01-27 03:31:42 EST
Please use accurate wording when filing a bug. NFS wasnt crashing anywhere in the tests that we conducted. The whole machine was freezing...two very different situations.
This bug is opened based on the ticket: http://support.gluster.com/rt/Ticket/Display.html?id=2504
The Gluster native NFS crashes when mounted locally over localhost and handling many parallel writes. 40 parallel "dd if=/dev/zero of=/nfsmount/inX.dat bs=1M" are started and the dd processes quickly hangs together with the NFS mount.
I could reproduce this issue inhouse while doing 40+ parallel dd's. The server itself get hanged while doing parallel writes in the mount point. Unable to run any commands after that and a reboot is required to bring the server back online.
More information available at 2251.
*** This bug has been marked as a duplicate of bug 2320 ***