Red Hat Bugzilla – Bug 216774
clusvcadm relocating virtual ip fails without complaining
Last modified: 2009-04-16 18:36:27 EDT
Description of problem:
When I tried to relocate my virtual IP from node roth-02 to
roth-01, it responded with "success" but it didn't do it.
I believe there was a problem relocating the service, in which
case it should have given an error and said "failed" or else
started it elsewhere and told me where it started the service
like it did for RHEL3.
It would also be nice if it told me why it couldn't relocate the
service, even if it's something like "service is already running" or
One could also argue that rgmanager should recognize that the service
is already running elsewhere, assume the user wants to move the vip
and do the necessary stop/start sequence, or something.
Version-Release number of selected component (if applicable):
Recreated at RHEL4U4, but I want to make sure it gets fixed in RHEL5.
Steps to Reproduce:
[root@roth-02 /]# clusvcadm -e 10.15.84.250 -m roth-01
Member roth-01 trying to enable 10.15.84.250...success
[root@roth-02 /]# ip addr show | grep 250
inet 10.15.84.250/32 scope global eth1
(The service did not move and it still said "success" which misleads
Member roth-01 trying to enable 10.15.84.250...error
Service 10.15.84.250 could not be started on roth-01.
Reason: Service is already running on roth-02 (or some such)
Lon said he used to have a patch for this problem.
* The patch I had was in the case that you enabled a non-running service on a
node and the node failed to start it.
* If I read this correctly, you're using the '-e' flag to relocate an
already-running service, which won't work; try '-r' ?
I'll fix this today either way. '-e' will become '-r' if '-[nm]' is specified
and the service is already running.
ended up changing semantics to be:
-e reports an error (but returns success) if the service is already running.
Patches in CVS.
A package has been built which should help the problem described in
this bug report. This report is therefore being closed with a resolution
of CURRENTRELEASE. You may reopen this bug report if the solution does
not work for you.
Moving all RHCS ver 5 bugs to RHEL 5 so we can remove RHCS v5 which never existed.