Bug 226789

Summary: nfs failover does not work properly when specifying nfs_lock_workaround
Product: [Retired] Red Hat Cluster Suite Reporter: Josef Bacik <jbacik>
Component: clumanagerAssignee: Lon Hohberger <lhh>
Status: CLOSED ERRATA QA Contact: Cluster QE <mspqa-list>
Severity: medium Docs Contact:
Priority: medium    
Version: 3CC: cluster-maint, nobody+wcheng
Target Milestone: ---   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: RHBA-2007-0165 Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2007-06-14 12:53:47 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
patch to fix problem. none

Description Josef Bacik 2007-02-01 16:42:09 UTC
Description of problem:
If you specify a specific nfsclient, and you have nfs_lock_workaround 
specified, the nfs script will do an exportfs -u *:/dir, but that won't 
actually match properly and the mount will not be unexported.  This patch 
fixes the problem.

Comment 1 Josef Bacik 2007-02-01 16:42:09 UTC
Created attachment 147113 [details]
patch to fix problem.

Comment 2 Lon Hohberger 2007-02-01 17:02:02 UTC
The patch is just a reversion to 1.5 from CVS:

Index: src/services/svclib_nfs
===================================================================
RCS file: /cvs/devel/clumanager/src/services/svclib_nfs,v
retrieving revision 1.7
diff -u -r1.7 svclib_nfs
--- src/services/svclib_nfs     16 Jun 2006 16:45:20 -0000      1.7
+++ src/services/svclib_nfs     1 Feb 2007 17:01:44 -0000
@@ -18,7 +18,7 @@
 #  MA 02139, USA.
 #
 
-# $Revision: 1.7 $
+# $Revision: 1.5 $
 #
 # Author: Tim Burke <tburke at redhat.com>
 
@@ -420,7 +420,6 @@
 
        typeset svc_name=$(getSvcName $DB $svcID) > /dev/null
        typeset export_directory=$(getSvcExportDir $DB $dirtoken) > /dev/null
-       typeset nfs_lock_workaround=$(getSvcMgrNFSLock $DB)
 
         case "$command" in
           start) 
@@ -441,14 +440,7 @@
                    typeset export_args="$export_opt $export_options
$export_client:$export_directory" 
                    ;;
           stop)     
-                   # Use exportfs -u * always to make sure it gets
-                   # sent to the kernel if the workaround is set
-                   if [ "$nfs_lock_workaround" = "yes" ] || \
-                      [ "$nfs_lock_workaround" = "1" ]; then
-                     typeset export_args="-u *:$export_directory" 
-                   else
-                     typeset export_args="-u $export_client:$export_directory" 
-                   fi
+                   typeset export_args="-u $export_client:$export_directory" 
                    ;;
          *)        
                    logAndPrint $LOG_ERR "export_client: Invalid command $command"


Comment 11 Lon Hohberger 2007-05-21 20:10:22 UTC
Tested failover with child-subdirectories (208995 / w/ respinned package) w/
nfs_lock_workaround=1 set.

* service restarted and relocated correctly even with locks on the file system
* rhel4 clients worked fine
* rhel5 client had a problem, but this is not a clumanager issue.


Comment 13 Red Hat Bugzilla 2007-06-14 12:53:47 UTC
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 the 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-2007-0165.html