Questions and Answers : Unix/Linux : Using CluBORun
Message board moderation

To post messages, you must log in.


Send message
Joined: 30 Oct 18
Posts: 16
Credit: 192,218,221
RAC: 0
Message 40179 - Posted: 17 Oct 2019, 10:53:20 UTC


I am trying to use BOINC in a cluster with restriction of use their own batch system, so I can not run BOINC in a parallel way (the people of the cluster wants that all the jobs get registered in the batch system).

I think that this can be done by using CluBORun, but there is not a lot of information about this software. Anyone know how CluBORun exactly works? I will thanks any other option.

Thank you!
ID: 40179 · Report as offensive     Reply Quote
Henry Nebrensky

Send message
Joined: 13 Jul 05
Posts: 162
Credit: 14,771,016
RAC: 113
Message 40180 - Posted: 17 Oct 2019, 12:02:51 UTC - in response to Message 40179.  

I saw you already asked about this in the Atlas forum. Could you run the BOINC tasks individually by submitting them as boinc_client --fetch_minimal_work --exit_after_finish to the batch system?
But I think you will still need persistent storage per-node for the BOINC directory, for any pending tasks, results etc..

Personally though, I feel that you're just making unnecessary work for yourselves by trying to mix two conflicting systems.

Sorry,I don't know anything about CluBORun.
ID: 40180 · Report as offensive     Reply Quote

Questions and Answers : Unix/Linux : Using CluBORun

©2023 CERN