[core] Move create call / receiving request off of execution io_context #55904
+55
−35
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Outdated suggestions cannot be applied.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.
Why are these changes needed?
Today our gRPC servers have polling thread(s). They end up not actually doing much work because the actual copying of the requests into the proto objects is offloaded to the main service io_context thread(s). When sending relatively large amounts of inlined object data (~1-5mb) we found that there can be a bottleneck on actually receiving the args because the args are copied into the proto on the single-threaded io_context that is also doing a lot of other work. The # of polling threads meanwhile can be scaled up without concurrency issues and they're much less busy.
This could also help offload work off the GCS's main io context + fully utilize the # of object manager threads.