ip-isolate flag for distributed pytorch on lassen

radical-cybertools / radical.saga

A Light-Weight Access Layer for Distributed Computing Infrastructure and Reference Implementation of the SAGA Python Language Bindings.

http://radical-cybertools.github.io/saga-python/

Other

83 stars 34 forks source link

ip-isolate flag for distributed pytorch on lassen #815

Closed lee212 closed 3 years ago

lee212 commented 3 years ago

This PR aims to introduce "-alloc_flags ipisolate" for lsf jobs on Lassen. One use case is that distributed pytorch requires sockets opened without authentication, which isn't abiding LC security policy. This flag is accommodating with the system, and for more details: https://lc.llnl.gov/confluence/pages/viewpage.action?pageId=650674651

mtitov commented 3 years ago

I would think that after switching to Config class, this could be moved into _verify method