[Linux-cluster] openais problem

[Steven Dake]

Which openais package version are you using?

If it is the latest, this indicates a failure in multicast on your
switch hardware.  What kind of switch hardware are you using?  You might
try using a dumb device switch and see if that fixes the problem.  You
could also try overriding the MTU of totem via flags to the
configuration file.  It is possible your switch will not pass fragmented
multicast packets which occur rarely after a configuration change.  I
have seen switches that are defective which wont pass fragmented
multicast packets, especially some early generation ATCA switch hardware
or softswitches.

vserver = virtual server?  What type...

If you have Cisco hardware review:

http://www.openais.org/doku.php?id=faq:cisco_switches

On Wed, 2008-02-13 at 09:19 -0500, Marcos Ferreira da Silva wrote:
> I have two node cluster.
> 
> I start node 1.
> When I start node2 this join to cluster and appear this at log messages:
> 
> Feb 13 11:13:27 vserver1 openais[28108]: [TOTEM] FAILED TO RECEIVE 
> Feb 13 11:13:27 vserver1 openais[28108]: [TOTEM] entering GATHER state from 6. 
> 
> [root at vserver2 ~]# clustat  
> Member Status: Quorate
> 
>   Member Name                        ID   Status
>   ------ ----                        ---- ------
>   vserver1.teste.br                    1 Offline
>   vserver2.teste.br                    2 Online, Local, rgmanager
> 
> The firewall is stopped.
> 
> [root at vserver2 ~]# cat /etc/hosts
> # Do not remove the following line, or various programs
> # that require network functionality will fail.
> 127.0.0.1               localhost.localdomain localhost
> ::1                     localhost6.localdomain6 localhost6
> 192.168.200.200         vserver1.teste.br teste1
> 192.168.200.201         vserver2.teste.br teste2
> 
> My cluster.conf:
> 
> <?xml version="1.0"?>
> <cluster config_version="47" name="cluster1">
>         <fence_daemon clean_start="0" post_fail_delay="0" post_join_delay="30"/>
>         <fence_xvmd family="ipv4" key_file="/etc/cluster/fence_xvm.key" auth="none" hash="none"  />
>         <clusternodes>
>                 <clusternode name="vserver1.teset.br" nodeid="1" votes="1">
>                         <fence>
>                                 <method name="1">
>                                         <device domain="admin" name="xvm"/>
>                                 </method>
>                         </fence>
>                 </clusternode>
>                 <clusternode name="vserver2.teset.br" nodeid="2" votes="1">
>                         <fence>
>                                 <method name="1">
>                                         <device domain="intranetteste" name="xvm"/>
>                                 </method>
>                         </fence>
>                 </clusternode>
>         </clusternodes>
>         <cman expected_votes="1" two_node="1"/>
>         <fencedevices>
>                 <fencedevice agent="fence_xvm" key_file="/etc/cluster/fence_xvm.key" auth="none" hash="none" name="xvm"/>
>         </fencedevices>
>         <rm>
>                 <failoverdomains/>
>                 <resources/>
>     <vm autostart="0" name="admin" path="/etc/xen"/>
>     <vm autostart="0" name="intranetteste" path="/etc/xen"/>
>         </rm>
>         <totem token="21000"/>
> </cluster>
> 
> ------------------------------
> Marcos Ferreira da Silva
> Digital Tecnologia
> Uberlandia-MG
> (34) 9154-0150 / 3226-2534
> 
> 
> 
> --
> Linux-cluster mailing list
> Linux-cluster at redhat.com
> https://www.redhat.com/mailman/listinfo/linux-cluster