[Linux-HA] Tomcat times out

[Andrew Beekhof]

On Mon, Jul 14, 2008 at 08:34, Ehlers, Kolja <ehlers at clinresearch.com> wrote:
> Hello everybody,
>
> sometimes in my cluster it happens that tomcat the tomcat start or monitor
> does not return SUCCESS inside the timeout frame. Then I get:
>
> Failed actions:
>    tomcat_22:1_start_0 (node=www2test, call=25, rc=-2): Timed Out
>    tomcat_22sdb:1_start_0 (node=www2test, call=23, rc=-2): Timed Out
>    tomcat_34:1_start_0 (node=www2test, call=30, rc=-2): Timed Out
>    tomcat_30:1_start_0 (node=www2test, call=28, rc=-2): Timed Out
>
> Can I somehow learn heartbeat that this error code is not fatal and that it
> can retry this resource again. It just cleans it up totally and never tries
> to start again.

better yet, tell heartbeat to wait long enough so that the actions
don't time out.
Look up the default-action-timeout option.

>
> Thanks
>
> -----Ursprungliche Nachricht-----
> Von: linux-ha-bounces at lists.linux-ha.org
> [mailto:linux-ha-bounces at lists.linux-ha.org]Im Auftrag von Michael Alger
> Gesendet: Montag, 7. Juli 2008 17:15
> An: General Linux-HA mailing list
> Betreff: Re: [Linux-HA] Tomcat times out
>
>
> On Mon, Jul 07, 2008 at 04:18:48PM +0200, Ehlers, Kolja wrote:
>>> I am running Apache and one virtual ip adress in one group now
>>> and 9 tomcat instances ungrouped. I want heartbeat to start and
>>> monitor the apache/ip group on only one node and all tomcats on
>>> both nodes. So if apache or the ip fails merge the ip and apache
>>> to the other node, the tomcats are already started there.
>>
>> what I probably could do to have my 9 tomcat instances on both
>> nodes is create another 9 resources. Then apply to the first ones
>> a rule to only run on node1 and to the other to stay on node2. Is
>> this the proper way to do this?
>>
>> For explanation: I will be using the apache_jk to balance the load
>> to the different tomcat workers and this will only work if all
>> tomcats run on both nodes.
>
> It may work, I haven't tried an active/active setup and looking at the
> example I'm not entirely sure why it lets you specify two primitives
> with the same id ("apache"). I suppose it must be because they're in
> different groups?
>
> Anyway, you might want to read about clones:
>
>  http://www.linux-ha.org/v2/Concepts/Clones
>
> I'm not sure how you provide different configuration information to
> each clone. My best guess is you'd keep your existing tomcat resource
> definitions, but have heartbeat run two clones of each one (limiting
> each node to run a single clone of each resource).
>
> It sounds like the tomcat OCF script is pretty functional, so this
> should actually be pretty easy. Since you don't seem to want to make
> use of the state of the tomcat resources in any way, this is probably
> all you need to do to get exactly what you want.
> _______________________________________________
> Linux-HA mailing list
> Linux-HA at lists.linux-ha.org
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>
> Geschäftsführung: Dr. Michael Fischer, Reinhard Eisebitt
> Amtsgericht Köln HRB 32356
> Steuer-Nr.: 217/5717/0536
> Ust.Id.-Nr.: DE 204051920
> --
> This email transmission and any documents, files or previous email
> messages attached to it may contain information that is confidential or
> legally privileged. If you are not the intended recipient or a person
> responsible for delivering this transmission to the intended recipient,
> you are hereby notified that any disclosure, copying, printing,
> distribution or use of this transmission is strictly prohibited. If you
> have received this transmission in error, please immediately notify the
> sender by telephone or return email and delete the original transmission
> and its attachments without reading or saving in any manner.
>
> _______________________________________________
> Linux-HA mailing list
> Linux-HA at lists.linux-ha.org
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>