I plan to attend the FTWG on Tuesday. I agree with the general direction for the next steps. I believe that what we discussed was to enable basic mechanisms for failure detection, (and possibly some form of failure notification) and to leave recovery mechanisms for later. Am I right? Should we set some time at the FTWG to discuss this? Ignacio On 11/27/18 11:31 AM, Bland, Wesley via mpiwg-ft wrote:

Hi all,

Unfortunately, I have a conflict that’s come up tomorrow that I can’t move around so I’ll be unable to attend the FTWG call. I’m not sure if there’s much for us to discuss unless Aurelien has something new about the Init/Finalize PR. I think that’s pretty much locked down by the 2 week deadline though, so if there are no objections, I propose we cancel and meet up next week for the F2F meeting in California. The FTWG time is Tuesday afternoon so hopefully everyone can make it.

For now, I think the plan will be to work on the next FT project which is deciding what to do with the rest of the ULFM proposal. I believe the next thing we’d previously planned was to work on the generic FT definitions and specifications in the “intro” of the chapter and leave the new API functions for the next step after that.

Thanks, Wesley _______________________________________________ mpiwg-ft mailing list [email protected] https://lists.mpi-forum.org/mailman/listinfo/mpiwg-ft