wang-mask
wang-mask
@odidev , I've reproduced the issue. It`s because the shortage of memory (in the doc, the runtime needs 4GiB mem). Maybe you can reduce the runtime memory or get some...
I think the launcher pod should be created after all the workers are ready when the `spec.launcherCreationPolicy` is set to `"WaitForWorkersReady"`, even if the MPIJob is suspend. The current logic...
> I think that when the MPIJob is suspended, workers aren't created by the mpi-operator. Could you clarify such a situation? In this case the workers are not created by...
My understanding of `spec.launcherCreationPolicy is set to "WaitForWorkersReady"` is to wait for all workers to be ready before creating a launcher. So in a suspended state, the workers have not...
> > > I think that when the MPIJob is suspended, workers aren't created by the mpi-operator. Could you clarify such a situation? > > > > > > In...
> With this implementation: what happens if the job is running, then it is suspended and unsuspended? > > Is a launcher pod created as soon as it is unsuspended...
It is the same failure, maybe it is time to update the example.