Re: [Condor-users] high availability question

[Nick LeRoy <nleroy@xxxxxxxxxxx>]

On Tue September 9 2008, Justin Ottley wrote:
> hey all,
Hello Justin,

> im trying to confirm whether the high availability behavior i've observed
> is expected -
>
> I have a central manager high availability setup going, following the
> example configuration (
> http://www.cs.wisc.edu/condor/manual/v7.1/3_10High_Availability.html
> section 3.10.2.4). On both the active central manager and backup central
> manager i can do condor_status (in particular condor_status -any) to get
> the current cluster state. I can also run condor_status -any on the (one)
> other machine in the cluster (scheduler). condor_status shows the HAD
> daemons (and backup machine daemons).
>
> When i disconnect the first central manager, the second picks up the
> negotiator, and the user priorities, so no problems there. At this point
> when I run condor_status -any the command blocks unless i specify the 2nd
> central manager explicitly (via -pool). Same goes for condor_submit and
> condor_q. When the first central manager comes back, it all works fine
> again (without -pool). Is this the expected behavior?

Do you have a collector running on both machines?  Are both listed in the 
COLLECTOR_HOST configuration?

-Nick

-- 
           <<< Follow the white rabbit. >>>
 /`-_    Nicholas R. LeRoy               The Condor Project
{     }/ http://www.cs.wisc.edu/~nleroy  http://www.cs.wisc.edu/condor
 \    /  nleroy@xxxxxxxxxxx              The University of Wisconsin
 |_*_|   608-265-5761                    Department of Computer Sciences