Fix Streaming Issues & Cleanup Async #493
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
bolasim
approved these changes
Aug 1, 2023
| await queue.put(ResponseChunk(chunk)) | ||
| except Exception as e: | ||
| self._logger.exception("Exception while reading stream response: " + str(e)) | ||
| finally: |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
Summary
I noticed a couple issues while working through the Streaming stuff.
There is an issue with how we were handling semaphore management. We were offloading the user's predict work to a background task, and releasing the semaphore before that finishes. This means that users might be getting more concurrency than they ask for.
If a stream throws an error partway through, right now, the user does not see a good error in the logs, and the client hangs for a minute.
In addition, in this PR, did some cleanup on the async stuff.
Solutions
None, so that the response generator can terminate, and doesn't hang for 60 seconds.Testing
Added integration tests for both of these complicated cases, and will test manually as well on dev.