On Tuesday 08 July 2008, codeline wrote:
> Sorry pingd is bad example.  The real question is that if you had  resource
> name foo and it fails on node1. How would get foo to restart on node2
> instead of the original node1?

set the 'resource-failure-stickiness' to an appropriate value .... see  
http://www.linux-ha.org/ScoreCalculation

Regards,
Andreas

>
> Thanks
>
> Ivan-58 wrote:
> > Hi,
> >
> > Not sure if I understood what you are after but see this page:
> >
> > http://www.linux-ha.org/v2/faq/pingd
> >
> > and section:
> > "Only Run my_resource on Nodes with Access to at Least One Ping Node"
> >
> > Ivan
> >
> > On Tue, 2008-07-08 at 10:16 -0700, codeline wrote:
> >> Hello,
> >>
> >> I have added the following resource and constraints.
> >>
> >> <resources>
> >>         <group id="ping_test_group">
> >>                 <primitive id="ping_test" class="ocf"
> >> provider="heartbeat" type="pingd">
> >>                         <operations>
> >>                                 <op id="ping_mon" interval="20s"
> >> name="monitor" timeout="60s"/>
> >>                         </operations>
> >>                 </primitive>
> >>         </group>
> >> </resources>
> >> <constraints>
> >>         <rsc_location id="rsc_test_group" rsc="ping_test_group">
> >>                 <rule id="rsc_location_group_1" score="100">
> >>                         <expression attribute="#uname" operation="eq"
> >> value="node10"/>
> >>                 </rule>
> >>         </rsc_location>
> >> </constraints>
> >>
> >> When I kill pingd on node10, it will restart it on node10. I want the
> >> resource to move over to the next node if it fails. Is there a way of
> >> doing this?
> >>
> >> Thanks for the help.
> >
> > _______________________________________________
> > Linux-HA mailing list
> > [email protected]
> > http://lists.linux-ha.org/mailman/listinfo/linux-ha
> > See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to