Description of problem: When a node owns any plock resources on a file system and that fs is unmounted, the remaining nodes do nothing to change the owner value on those resources. Any process that attempts to access those plock resources will become stuck and require a reboot. The fix is to change the owner to 0 (unowned) on any resources owned by a node that unmounts. This bug only applies when plock_ownership=1, which is not the default setting (and ownership mode has never really worked correctly.) Test: node1: mount /gfs; cd /gfs; lock_load node2: mount /gfs; cd /gfs; lock_load node3: mount /gfs; cd /gfs; lock_load node1: kill lock_load; cd; umount /gfs nodes 2&3: lock_load permanently hangs Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info:
Created attachment 439017 [details] patch fix
pushed to RHEL56 branch http://git.fedorahosted.org/git?p=cluster.git;a=commitdiff;h=e8cc459b6ac847c1dd92a17ce833b74e46228ab0
An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on therefore solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHBA-2011-0036.html