flux-framework
Parallel debugging description
As discussed this at 11AM meeting 7/16, we need a high level text on how to parallel debug a flux job. A good time will be after Perforce Software's support engineer will have a chance to poke at a Flux version on one of LC TOSS clusters. I will coordinate.
In the meanwhile, should this text go into https://flux-framework.readthedocs.io/en/latest/quickstart.html or somewhere else?
In the meanwhile, should this text go into https://flux-framework.readthedocs.io/en/latest/quickstart.html or somewhere else?
Yeah, I think it seems reasonable to have a parallel debugging section on the Quickstart page for now. Eventually, maybe we can piece that section out and have a separate page just on debugging?
Eventually, maybe we can piece that section out and have a separate page just on debugging?
Debugging and performance tuning, and yes eventually :-)
https://github.com/flux-framework/flux-sched/pull/694#issuecomment-663890636
I will be seeking some testing from Perforce and document in the next few weeks.
As part of this testing, some more best-practice recommendations as to how to configure the TotalView parallel debugger will be likely come out. That should be documented too.