Re: [Condor-users] MPI condor Config

["Diego Bello" <dbello@xxxxxxxxx>]

On 10/28/06, Diego Bello <dbello@xxxxxxxxx> wrote:
> On 10/24/06, Becky Gietzel <bgietzel@xxxxxxxxxxx> wrote:
> >
> > On Oct 20, 2006, at 5:50 PM, Diego Bello wrote:
> >
> > > Hi everyone, I have a Condor pool made of workstations to support MPI,
> > > simple jobs and dag, all using globus. Condor version is 6.8.0
> > >
> > > What I need is that MPI jobs could be stopped if a machine is used,
> > > wich normally is between 10 am and 9 pm. I have tried some
> > > configurations taken from the condor manual, but some jobs doesn't
> > > start. I think there could be a problem with the start configuration.
> > >
> > > I'm now trying a dag job, with three jobs doing nothing more than
> > > /bin/hostname, but it gets to the queue, the first job starts running
> > > but, after several hours, it doesn't finish. If I send a globus job
> > > directly, it works. My proxy is valid for 48 hrs.
> > >
> > > I have attached my central manager and my exec nodes's config files.
> > > Can someone tell me if there is something wrong with my config files?.
> > >
> >
> > You'll want to adjust your START policy for the execute nodes as
> > follows:
> >
> > Add:
> > IsNighttime = (ClockMin < 600 || ClockMin > 1260)
> >
> >
> > Replace the START and PREEMPT expressions with:
> > START   = ( (Scheduler =?= $(DedicatedScheduler) && $(IsNighttime) =?
> > = TRUE && $(KeyboardIdleTime) > $(StartIdleTime) ) ||  $(START) )
> > PREEMPT = (Scheduler =!= $(DedicatedScheduler) && $(KeyboardBusy)
> >
> > This policy will allow MPI jobs to start only during the nighttime
> > hours if nobody is actively using the machine.  Once you set up the
> > new policy, make sure you are able to run a simple Vanilla universe /
> > bin/hostname job.  When that is working try the dag job with /bin/
> > hostname again.  Then try an MPI job.
> >
> > If you are using the MPI Universe for your MPI jobs I'd recommend
> > switching to the Parallel Universe.
> >
> >
> > Thanks,
> >
> > Becky
> >
>
> Thanks for the reply!
>
> I tried what you said, but condor daemons can't start in exec nodes.
> This is the error message i get:
>
> *** Last 20 line(s) of file StartLog:
> 10/28 23:00:39 Using config source: /etc/condor/condor_config
> 10/28 23:00:39 Using local config sources:
> 10/28 23:00:39    /opt/condor-6.8.0/local.chaparro/condor_config.local
> 10/28 23:00:39 DaemonCore: Command Socket at <200.1.19.171:9642>
> 10/28 23:00:39 ERROR "Syntax error in START expression: '( (Scheduler
> =?= "DedicatedScheduler@xxxxxxxxxxxxxxxxxxx" && (ClockMin < 600 ||
> ClockMin > 1260) =?= TRUE &&  > 15 * 60 ) ||  ( (KeyboardIdle > 15 *
> 60) && ( ((LoadAvg - CondorLoadAvg) <= 0.3) || (State != "Unclaimed"
> && State != "Owner")) ) )'" at line 286 in file util.C
> 10/28 23:00:52 passwd_cache::cache_uid(): getpwnam("condor") failed:
> user not found
> 10/28 23:00:52 passwd_cache::cache_uid(): getpwnam("condor") failed:
> user not found
> 10/28 23:00:52 ******************************************************
> 10/28 23:00:52 ** condor_startd (CONDOR_STARTD) STARTING UP
> 10/28 23:00:52 ** /opt/condor-6.8.0/sbin/condor_startd
> 10/28 23:00:52 ** $CondorVersion: 6.8.0 Jul 19 2006 $
> 10/28 23:00:52 ** $CondorPlatform: I386-LINUX_RHEL3 $
> 10/28 23:00:52 ** PID = 5351
> 10/28 23:00:52 ** Log last touched 10/28 23:00:39
> 10/28 23:00:52 ******************************************************
> 10/28 23:00:52 Using config source: /etc/condor/condor_config
> 10/28 23:00:52 Using local config sources:
> 10/28 23:00:52    /opt/condor-6.8.0/local.chaparro/condor_config.local
> 10/28 23:00:52 DaemonCore: Command Socket at <200.1.19.171:9683>
> 10/28 23:00:52 ERROR "Syntax error in START expression: '( (Scheduler
> =?= "DedicatedScheduler@xxxxxxxxxxxxxxxxxxx" && (ClockMin < 600 ||
> ClockMin > 1260) =?= TRUE &&  > 15 * 60 ) ||  ( (KeyboardIdle > 15 *
> 60) && ( ((LoadAvg - CondorLoadAvg) <= 0.3) || (State != "Unclaimed"
> && State != "Owner")) ) )'" at line 286 in file util.C
> - Hide quoted text -
> *** End of file StartLog
>
>
> I tried having a START=TRUE expression before what you said, and then
> removing that. The only difference was the error message saying TRUE
> instead ((LoadAvg - Condor.......
>
> In the PREEMT line, I supose there is a missing ")" at the end, am I right?.
>
> Can you help me find out what is going wrong?
>
> Thanks.
>
> --
> Diego Bello Carreño
> Estudiante Memorista de Ingeniería Civil Informática
> UTFSM, Valparaíso, Chile
> Usuario #294897 counter.li.org

I have finally found the reason!!!

I though that "KeyboardIdleTime" was some condor variable, but NO!, I
have to define it. So I gave it a value of 60 (mins) and now it
works!!!

Thanks very much Becky for your help.

--
Diego Bello Carreño
Estudiante Memorista de Ingeniería Civil Informática
UTFSM, Valparaíso, Chile
Usuario #294897 counter.li.org