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

Re: [Condor-users] Question about using Condor in a non-Grid environment

Hi Jan,

Your case is somewhat similar to MPI/Parallel Universe but different.
Similar in the sense that you have to allocate all the machines (or) none.
For example, if the client machines are available and not the server, then none of the client applications would work.
So, there is actually a dependency. DAGMAN comes to my mind.
The next question is -- Are the client machines inter-dependent too? OR Do they depend only on the server?
If they are dependent on each other then it is a cyclic dependency - that you cannot handle with DAGMAN (Experts, Feel free to correct me here.)
What you need is a custom resource allocator - which allocates the machine separately and then instruct condor to run the jobs that you would like to.
I am not sure if this is possible with Condor. I too had wished for this feature but could not find

I am also a starter here. Condor could still work. Good Luck!

Best Regards,

-----Original Message-----
From: condor-users-bounces@xxxxxxxxxxx [mailto:condor-users-bounces@xxxxxxxxxxx] On Behalf Of Jan_Seidel@xxxxxxxx
Sent: Friday, July 20, 2012 7:13 AM
To: condor-users@xxxxxxxxxxx
Subject: Re: [Condor-users] Question about using Condor in a non-Grid environment

From: condor-users-bounces@xxxxxxxxxxx [mailto:condor-users-bounces@xxxxxxxxxxx] On Behalf Of Rich Pieri
> If I follow your description, a submission to the batch system would 
> entail a four-way dependency between jobs, each of which runs 
> concurrently on a specific type of node.  Two of these jobs function as 
> servers and the other two function as clients.  Concurrent, 
> heterogeneous batch scheduling.

I'm not sure if  that's the correct description, but maybe I'm not following you correctly. Let me rephrase the problem:
We have one control server from where we start all the jobs. The jobs communicate with all nodes by themselves (we have an SSH- and Telnet-based solution that's already working). The only thing we need a resource manager / job queue for is to manage the available resources and kick off the jobs once the required resources are available.

User 1 submits a job which needs 1 server, 2 linux clients, 1 windows client
User 2 submits a job which needs 2 servers, 2 linux clients, 2 windows clients

Condor in this case should only manage the resources and kick off the jobs once the required nodes are available. I'm not sure if that's a typical use case of Condor, though.


Condor-users mailing list
To unsubscribe, send a message to condor-users-request@xxxxxxxxxxx with a
subject: Unsubscribe
You can also unsubscribe by visiting

The archives can be found at:


The contents of this e-mail and any attachment(s) are confidential and intended for the named recipient(s) only.
E-mail transmission is not guaranteed to be secure or error-free as information could be intercepted, corrupted,
lost, destroyed, arrive late or incomplete, or may contain viruses in transmission. The e mail and its contents
(with or without referred errors) shall therefore not attach any liability on the originator or HCL or its affiliates.
Views or opinions, if any, presented in this email are solely those of the author and may not necessarily reflect the
views or opinions of HCL or its affiliates. Any form of reproduction, dissemination, copying, disclosure, modification,
distribution and / or publication of this message without the prior written consent of authorized representative of
HCL is strictly prohibited. If you have received this email in error please delete it and notify the sender immediately.
Before opening any email and/or attachments, please check them for viruses and other defects.