Project

General

Profile

Idea #3980

Investigate if the MPI program can be run in such a way that the loss of a single process doesn't stop the whole program

Added by Kurt Biery over 6 years ago. Updated over 5 years ago.

Status:
Closed
Priority:
Normal
Assignee:
-
Target version:
-
Start date:
06/04/2013
Due date:
% Done:

0%

Estimated time:
Duration:

Description

The ability to keep running when a single EventBuilder or the online monitoring instance of the Aggregator dies (and probably end the run gracefully soon thereafter) would be quite useful.

We should investigate if this is possible from the MPI perspective and what changes would need to be made to the DAQ applications to support this. For example, would we need to change the BoardReader code to skip over a missing EB?
--Kurt


Related issues

Related to artdaq - Idea #5986: Investigate whether we can support graceful loss of a small number of EventBuildersClosed04/21/2014

History

#1 Updated by Kurt Biery over 5 years ago

  • Status changed from New to Closed

This Issue has been superseded by #5986.



Also available in: Atom PDF