[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Condor-users] Unable to process machine ClassAds
- Date: Thu, 23 Jul 2009 09:42:48 -0500
- From: Matthew Farrellee <matt@xxxxxxxxxx>
- Subject: Re: [Condor-users] Unable to process machine ClassAds
Nilson Cain wrote:
> Hi there.
> I'm trying to get some stuff set up with a Condor grid. I have a central
> manager which is also a compute host, with 4*x86_64 cores. I also have two
> 32-bit slots on a 32 bit SMP system, and two 32 slots on a virtualized
> (Pools of Virtual Boxes) install.
> I have all systems showing in the condor_status output.
> I'm running into problems when trying to dispatch test jobs (from examples/)
> to the pool. 64-bit jobs have run with success in the past, but 32 bit jobs
> are not dispatching. I'm pretty sure I am running into conflicts with the
> ClassAd requirements, but I can't view the classad that is generated by the
> examples/submit script. Running condor_q -analyze shows the jobs and the
> host from which they originated, but all are idle.
> It shows that machines were rejected by my job requirements, but I get
> "Unable to process machine ClassAds" in that output, so I can't verify what
> it is requesting.
> Anyone encountered this error before?
You should try condor_q -better-analyze. You can see the job
requirements with condor_q -l | grep Requirements, and machine with
condor_status -l | grep START