[Linux-cluster] Ricci doesn't work

[Chris Feist]

On 08/13/12 16:38, Chip Burke wrote:
> Ricci is seemingly not working through either Luci nor cman_tool. There does't
> seem to be a lot of logging to go off of (at least I haven't found it) but what
> I did find in the Luci log is as follows:
>
> 15:51:06,793 ERROR [luci.lib.ricci_communicator] Error receiving header from
> node2.domain.local:11111
> Traceback (most recent call last):
>    File "/usr/lib64/python2.6/site-packages/luci/lib/ricci_communicator.py",
> line 121, in __init__
>      hello = self.__receive(self.__timeout_init)
>    File "/usr/lib64/python2.6/site-packages/luci/lib/ricci_communicator.py",
> line 503, in __receive
>      errstr = _('Error reading from %s:%d: %s') \
>    File "/usr/lib/python2.6/site-packages/pylons/i18n/translation.py", line 106,
> in ugettext
>      return pylons.translator.ugettext(value)
>    File "/usr/lib/python2.6/site-packages/paste/registry.py", line 137, in
> __getattr__
>      return getattr(self._current_obj(), attr)
>    File "/usr/lib/python2.6/site-packages/paste/registry.py", line 197, in
> _current_obj
>      'thread' % self.____name__)
> TypeError: No object (name: translator) has been registered for this thread
> 15:51:06,793 ERROR [luci.lib.ricci_helpers] Error receiving header from
> node2.XXXX.local:11111
> 15:51:06,793 ERROR [luci.lib.ricci_helpers] Error retrieving batch number from
> node3.XXXXX.local: Error receiving header from node3.XXXXX.local:11111
>
> Cluster config I am trying to push:
>
> <?xml version="1.0"?>
> <cluster config_version="27" name="Xanadu">
> <clusternodes>
> <clusternode name="xanadunode1" nodeid="1">
> <fence>
> <method name="Method">
> <device name="VMWare_Fence" port="XanaduNode1" ssl="on"/>
> </method>
> </fence>
> </clusternode>
> <clusternode name="xanadunode2" nodeid="2">
> <fence>
> <method name="Method">
> <device name="VMWare_Fence" port="XanaduNode2" ssl="on"/>
> </method>
> </fence>
> </clusternode>
> <clusternode name="xanadunode3" nodeid="3">
> <fence>
> <method name="Method">
> <device name="VMWare_Fence" port="XanaduNode3" ssl="on"/>
> </method>
> </fence>
> </clusternode>
> </clusternodes>
> <cman expected_votes="6"/>
> <fencedevices>
> <fencedevice agent="fence_vmware_soap" ipaddr="vsphere.XXXXXX.local"
> login="vmwarefence" name="VMWare_Fence" passwd="XXXXXXXX"/>
> </fencedevices>
> <quorumd label="quorum" votes="3"/>
> </cluster>
>
>
> Running config:
>
> <?xml version="1.0"?>
> <cluster config_version="26" name="Xanadu">
> <clusternodes>
> <clusternode name="xanadunode1" nodeid="1">
> <fence>
> <method name="Method">
> <device name="VMWare_Fence" port="XanaduNode1" ssl="on"/>
> </method>
> </fence>
> </clusternode>
> <clusternode name="xanadunode2" nodeid="2">
> <fence>
> <method name="Method">
> <device name="VMWare_Fence" port="XanaduNode2" ssl="on"/>
> </method>
> </fence>
> </clusternode>
> <clusternode name="xanadunode3" nodeid="3">
> <fence>
> <method name="Method">
> <device name="VMWare_Fence" port="XanaduNode3" ssl="on"/>
> </method>
> </fence>
> </clusternode>
> </clusternodes>
> <cman expected_votes="5"/>
> <fencedevices>
> <fencedevice agent="fence_vmware_soap" ipaddr="vsphere.XXXXX.local"
> login="vmwarefence" name="VMWare_Fence" passwd="XXXXXX"/>
> </fencedevices>
> <quorumd label="quorum" votes="2"/>
> </cluster>
>
> Any ideas? SCP and reboots are fun and all, but I would love Ricci to work.

Can you try using ccs to get the current configuration of that node:
ccs -h <host name> --getconf


As well as use ccs to try and set the conf on that node?
ccs -f <cluster.conf file> -h <host name> --setconf

This should let us narrow down whether it's an issue with ricci or luci.

Thanks!
Chris

>
> Thanks!
>
>
>
>
>
>
> --
> Linux-cluster mailing list
> Linux-cluster at redhat.com
> https://www.redhat.com/mailman/listinfo/linux-cluster
>