[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Condor-users] adding another submit host

eh, this does not work.

condor_q -direct schedd on the second submit host does not work. Even
before (naturally) and after the rsync.

On my second submit host I have the following running:
condor_procd  (with a lock?)..i am guessing this is normal.

One thing I can do is submit jobs using the second host, but my
primary host does not see the job submitted with this host. How can I
make that appear?


On Mon, Mar 22, 2010 at 6:50 PM, Mag Gam <magawake@xxxxxxxxx> wrote:
> Will try and let you know.
> I think the HAD is too complicated to setup therefore there should be
> an alternative to setup another submit host similar to Sun Grid
> Engine.
> On Mon, Mar 22, 2010 at 2:59 PM, Matthew Farrellee <matt@xxxxxxxxxx> wrote:
>> On 03/21/2010 09:49 AM, Mag Gam wrote:
>>> I currently have only 1 submit host in my pool. I would like to add
>>> another one. Instead of using the High Availability daemon
>>> (http://www.cs.wisc.edu/condor/manual/v7.2/3_1Introduction.html#9351)
>>> I would like to simply rsync the spool directory over. I will ofcourse
>>> run another collector on this second submit host and when the
>>> production negotiator is down I can manually start up the negotiator
>>> on this server. Would this work? Has anyone been doing this?
>> Never tried, but if you are doing manual fail-over I don't see why rsync wouldn't work. As long as your rsync is atomic for the entire SPOOL contents.
>> Best,
>> matt