[Difx-users] Infiniband device problems with Slurm/mpifxcorr
Joe Skeens
jskeens1 at utexas.edu
Wed Aug 18 15:43:30 EDT 2021
Hi all,
Thanks a lot for all the advice. I've started using mpispeed rather than
mpifxcorr to probe these issues, and I tried adding the "--mca btl
self,openib" tag to my commands, but I'm still having issues with
communication between nodes. It looks like there may be a problem with the
local setup of the cluster I'm working on (perhaps relating to the routing
table setup Walter pointed out). It looks like I need to do some more
digging locally.
Best,
Joe
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listmgr.nrao.edu/pipermail/difx-users/attachments/20210818/42bd1924/attachment.html>
More information about the Difx-users
mailing list