[0.9.1][Perf] Launch load kv task asynchronously with thread pool. #1612
+76
−18
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Outdated suggestions cannot be applied.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.
What this PR does / why we need it?
The Implementation of current
LLMDataDistCMgrConnector
connect and pull kv in synchronous manager which may brought drop of latency to the decode task if the consumer node consistently receiving tasks pushed from the remote producer node. The omni_infer launches llmdatadist'spull_kv
method in another thread which brings the better overlap between pull kv cache and model run. This implementation gains better performance against the synchronous path.In this PR, we bring this asynchronous philosophy into the vllm-ascend, and launch the
link
,pull_kv
andrequest_finished
tasks also in async managers.Does this PR introduce any user-facing change?
No any user interface change.
How was this patch tested?