dejanmm at fastmail
Nov 5, 2007, 5:51 AM
Post #7 of 7
(346 views)
Permalink
|
Hi,
On Fri, Nov 02, 2007 at 04:00:13PM +0000, Wojciech Turek wrote:
> Hi,
> On 2 Nov 2007, at 15:53, Dejan Muhamedagic wrote:
>
> >Hi,
> >
> >On Fri, Nov 02, 2007 at 02:52:28PM +0000, Wojciech Turek wrote:
> >>Hi,
> >>On 2 Nov 2007, at 13:38, Andrew Beekhof wrote:
> >>
> >>>On 11/2/07, Dejan Muhamedagic <dejanmm [at] fastmail> wrote:
> >>>>Hi,
> >>>>
> >>>>On Thu, Nov 01, 2007 at 04:21:41PM +0000, Wojciech Turek wrote:
> >>>>>Hi All,
> >>>>>
> >>>>>I have probably a two silly questions.
> >>>>>
> >>>>>1 How to stop resources (FileSystem resource) without stopping
> >>>>>HeartBeat?
> >>>>
> >>>>Set the target_role attribute on the resource to Stopped:
> >>>>
> >>>>crm_resource -r myrsc -p target_role -v Stopped
> >>>
> >>>this wont work in v1 :-)
> >>I thought so :-)
> >
> >Somehow missed the subject.
> >
> >>>>>2 If I stop HeartBeat on one node will it transfer resources to
> >>>>>it's
> >>>>>cluster partner or it will simply stop all resources on the node?
> >>>>
> >>>>There should be a failover, i.e. all resources should migrate to
> >>>>another node.
> >>How to stop resources without triggering fail-over?
> >
> >Just stop them.
> >
> >>Could I just
> >>umount Filesystem resources manually and then stop heartbeat service?
> >
> >Yes, but if there's another node alive it will start the
> >resources.
> >
> >>In general I am looking for clean and simple way to shutdown
> >>heartbeat and remove all the resources maintained by it.
> >
> >I don't understand what do you want to achieve. Shuting down
> >heartbeat on all nodes will also stop all resources. If you don't
> >want failover to happen, just first stop heartbeat on nodes which
> >don't run any services.
> I am sorry I wasn't clear enough. Both nodes are running resources
> from the shared storage so it is active-active fail-over. So you say
> if I stop heartbeat service on both nodes simultaneously it will stop
> resources without trying to give-up resources to another node and
> without shutting each-other in the head, right?
I'm afraid that you're out of luck. Your best bet is to stop all
resource manually and then try to kill heartbeats at the same
time. Perhaps using the KILL (9) signal.
Thanks,
Dejan
> Thanks
>
> Wojciech
>
> >
> >Thanks,
> >
> >Dejan
> >
> >
> >>Thanks,
> >>
> >>Wojciech
> >>>>
> >>>>Thanks,
> >>>>
> >>>>Dejan
> >>>>
> >>>>>Cheers,
> >>>>>
> >>>>>Wojciech turek
> >>>>>
> >>>>>
> >>>>>_______________________________________________
> >>>>>Linux-HA mailing list
> >>>>>Linux-HA [at] lists
> >>>>>http://lists.linux-ha.org/mailman/listinfo/linux-ha
> >>>>>See also: http://linux-ha.org/ReportingProblems
> >>>>_______________________________________________
> >>>>Linux-HA mailing list
> >>>>Linux-HA [at] lists
> >>>>http://lists.linux-ha.org/mailman/listinfo/linux-ha
> >>>>See also: http://linux-ha.org/ReportingProblems
> >>>>
> >>>_______________________________________________
> >>>Linux-HA mailing list
> >>>Linux-HA [at] lists
> >>>http://lists.linux-ha.org/mailman/listinfo/linux-ha
> >>>See also: http://linux-ha.org/ReportingProblems
> >>
> >>Mr Wojciech Turek
> >>Assistant System Manager
> >>University of Cambridge
> >>High Performance Computing service
> >>email: wjt27 [at] cam
> >>tel. +441223763517
> >>
> >>
> >>
> >>_______________________________________________
> >>Linux-HA mailing list
> >>Linux-HA [at] lists
> >>http://lists.linux-ha.org/mailman/listinfo/linux-ha
> >>See also: http://linux-ha.org/ReportingProblems
> >_______________________________________________
> >Linux-HA mailing list
> >Linux-HA [at] lists
> >http://lists.linux-ha.org/mailman/listinfo/linux-ha
> >See also: http://linux-ha.org/ReportingProblems
>
> Mr Wojciech Turek
> Assistant System Manager
> University of Cambridge
> High Performance Computing service
> email: wjt27 [at] cam
> tel. +441223763517
>
>
>
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
|
