Mailing List Archives
Authenticated access
|
|
|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Condor-users] clarification on job submit
- Date: Tue, 22 May 2007 15:34:27 +0100
- From: "Matt Hope" <matthew.hope@xxxxxxxxx>
- Subject: Re: [Condor-users] clarification on job submit
On 5/22/07, Partha sarathi <jinka.partha@xxxxxxxxx> wrote:
when i do a submit on condor i got a message
./condor_submit vanilla.submit
submitting job(s) .
1 job (s) submitted to cluster 8
what is this cluster 8 means???? I don't have put any name like cluster 8 on
my pool
This is auto generated.
All jobs in condor exist as a 'process' within a 'cluster' *, the
cluster is basically per submission (technically a cluster shares a
single executable) where as a job is per 'queue' statement in the
submission.
Since you are doing a simple submit you have a cluster with one process
There are a variety of performance benefits to using clusters when the
number of processes within the cluster increases.
If you run condor_q after submitting you will see a list of all the
jobs on that schedd. they will be identified by a composite number
such as 8.0.
This is ClusterId.ProcessId. If you wish to efficiently refer to a
particular job when talking to the queue (perhaps to hold it or
release it) you refer to it either as '8.0' or to refer to the cluster
as a whole just '8'
This identifier is not unique to the pool as a whole, just the schedd
(and it may be reused if you ever wipe the job queue for example).
* These are not the process and cluster names you would expect.
Process just means a single condor job. Cluster is simply a collection
of them.
Matt