[Linux-cluster] cluster.conf totem argument problems

[Digimer]

I've submitted this issue as a bug:

https://bugzilla.redhat.com/show_bug.cgi?id=614697

Cheers,

Madi

On 10-07-14 12:12 PM, Digimer wrote:
> Hi all,
>
> I'm trying to migrate my corosync.conf file's settings into
> cluster.conf. I discussed this with people on IRC, but I wanted to start
> a thread as I think this is a bigger issue that would benefit from a
> consistent thread.
>
> As I understand, any arguments in cluster.conf should be fed to the
> corosync preconfig. From there, defaults should be applied and the whole
> shebang should go into corosync-objdb. However. when I put a ticket[1]
> in for a related problem, Lon mentioned that some arguments are not yet
> supported and thus do not have support in cluster.rng. These two
> statements seem to contradict each other. Can this be clarified?
>
> Next, I stripped out the questionable arguments and tried to load cman
> with the cluster.conf having the totem arguments. Specifically:
>
> <totem rpp_mode="passive">
> <interface ringnumber="0" bindnetaddr="10.0.1.0"
> mcastaddr="226.94.1.1" mcastport="5405"/>
> </totem>
>
> With the above argument in my cluster.conf, it passed validation but
> then failed to start with:
>
> Starting cluster:
> Checking Network Manager... [ OK ]
> Global setup... [ OK ]
> Loading kernel modules... [ OK ]
> Mounting configfs... [ OK ]
> Starting cman... corosync died: Error, reason code is 8
> [FAILED]
>
> In the log file, I saw this error:
>
> Jul 14 00:00:46 an-node01 corosync[2227]: [MAIN ] Corosync Cluster
> Engine ('1.2.3'): started and ready to provide service.
> Jul 14 00:00:46 an-node01 corosync[2227]: [MAIN ] Corosync built-in
> features: nss rdma
> Jul 14 00:00:46 an-node01 corosync[2227]: [MAIN ] Successfully read
> config from /etc/cluster/cluster.conf
> Jul 14 00:00:46 an-node01 corosync[2227]: [MAIN ] Successfully parsed
> cman config
> Jul 14 00:00:46 an-node01 corosync[2227]: [MAIN ] Successfully
> configured openais services to load
> Jul 14 00:00:46 an-node01 corosync[2227]: [MAIN ] parse error in config:
> No multicast port specified
> Jul 14 00:00:46 an-node01 corosync[2227]: [MAIN ] Corosync Cluster
> Engine exiting with status 8 at main.c:1430.
>
> I don't think this is a syntax problem. When I remove the <totem />
> section and start cman, it comes up fine. Further, there is a totem
> ring, which I assume the adapter created out of defaults. I can see this
> via corosync-objctl (sorry, I don't have the exact output at the
> moment). There are several 'totem.X' entries, of which, one is 'mcastport'.
>
> Finally, when I shutdown cman, setup my corosync.conf file and start it
> by itself, the configuration loads. In fact, there I can specify a
> redundant ring and corosync-objctl shows both rings.
>
> Any idea what's going on here? Should I submit a new bug?
>
> Thanks in advance!
>
> 1. https://bugzilla.redhat.com/show_bug.cgi?id=613880


-- 
Digimer
E-Mail:         linux at alteeve.com
AN!Whitepapers: http://alteeve.com
Node Assassin:  http://nodeassassin.org