Hello,
I have a 5 node cluster.
virtsrv1n1
virtsrv2n2
virtsrv3n3
virtsrv4n4
virtsrv5n5

From Luci I am unable to manage virtsrv3n3 machine. Luci is unable to
reboot it for example and
if I select the node properties it shows me no status for Cluster
Daemons for this specific node.
All the other nodes are fully manageable from luci.
from command line everything seems to work fine.

net-cluster @ Sat May 12 00:53:33 2012
Member Status: Quorate

Member Name ID Status
------ ---- ---- ------
virtsrv1n1.mydomain.org 1 Online, Local, rgmanager
virtsrv2n2.mydomain.org 2 Online, rgmanager
virtsrv3n3.mydomain.org 3 Online, rgmanager
virtsrv4n4.mydomain.org 4 Online, rgmanager
virtsrv5n5.mydomain.org 5 Online, rgmanager

Looking the luci log i see this error if i try for exmaple to reboot the
node from luci:

00:12:19,603 ERROR [luci.lib.ricci_helpers] Unable to retrieve the batch
number from virtsrv3n3

ricci is working ok on the node as well as cman rgmanager and gfs2, and
I do not have iptables active.
Any hints ?

thank you

Rick




_______________________________________________
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

>00:12:19,603 ERROR [luci.lib.ricci_helpers] Unable to retrieve the batch
>number from virtsrv3n3

Looks like that comes from:
./usr/lib64/python2.6/site-packages/luci/lib/ricci_helpers.py

Whats unfortunate is their are several functions that emit that. I presume you
could make them all unique if you actually don't know what the real issue is.

Why dont you pastebin a bigger chunk of that log?

jlc
_______________________________________________
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

On 5/12/12 4:33 AM, Joseph L. Casale wrote:
>> 00:12:19,603 ERROR [luci.lib.ricci_helpers] Unable to retrieve the batch
>> number from virtsrv3n3
> Looks like that comes from:
> ./usr/lib64/python2.6/site-packages/luci/lib/ricci_helpers.py
>
> Whats unfortunate is their are several functions that emit that. I presume you
> could make them all unique if you actually don't know what the real issue is.
>
> Why dont you pastebin a bigger chunk of that log?
>
> jlc
> _______________________________________________
> CentOS mailing list
> CentOS@centos.org
> http://lists.centos.org/mailman/listinfo/centos
the only difference between this node and the other is that I added it
to the cluster manually and using commnd line,
everything is working except luci malfunctioning for this host...
_______________________________________________
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos

>the only difference between this node and the other is that I added it
>to the cluster manually and using commnd line,
>everything is working except luci malfunctioning for this host...

Lets try this slowly

Whatever the reason (sky is blue? who knows) your third node causes
an excpetion that ricci_helpers.py is written to catch.

If it weren't a problem according to ricci, it wouldn't throw an exception.

I have not read that file line by line, and you haven't posted more log lines.

All the exceptions unfortunately throw the same string.

If you want to know what fault (according to ricci) is being thrown, as I said,
post more log lines or make the strings unique so you know 'what' fault is
being thrown.

Make sense?
jlc
_______________________________________________
CentOS mailing list
CentOS@centos.org
http://lists.centos.org/mailman/listinfo/centos