MPI process wait for Head actor to be ready before to start (simulat… #54
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Doreisa fails if the MPI processes start calling the head actor before the analytics starts it. Solution was so far to put some sleep, but this always fails at some point. This PR proposes a code change to avoid this issue. The head actor is created either by the simulation processes or the analytics, whoever starts first and then the mpi processes way for the proper initialization of the head actor by the simulation.
Tested on a single node config only.
Little uncertainty if the head actor is created on the expected head node. That should be the case as stated by the Ray doc (https://docs.ray.io/en/latest/ray-core/api/doc/ray.runtime_context.get_runtime_context.html). But would need proper testing to be sure that when created by a MPI process the head actor does not run on this process node.