[dashboard][perf] Move Job package uploading to another thread. #45282
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
…blocking Signed-off-by: Ruiyang Wang <[email protected]>
jjyao
approved these changes
May 13, 2024
dashboard/modules/job/job_head.py
Outdated
| upload_package_to_gcs(package_uri, await req.read()) | ||
| data = await req.read() | ||
| await self._upload_package_thread_pool_executor.submit( | ||
| upload_package_to_gcs, package_uri, data |
There was a problem hiding this comment.
I saw it's just gcs kv put. Is it easy to just implement an async version of it using gcs_aio_client?
There was a problem hiding this comment.
I tried that, but upload_package_to_gcs is also used by CLI (Job SDK) code in sync form so it's non trivial to change.
I find that job submission APIs can take 30s, and it's only possible because of the large uploads, so I audited the code and found it |
Signed-off-by: Ruiyang Wang <[email protected]>
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Outdated suggestions cannot be applied.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.
Package uploading is a CPU intensive work in Dashboard, where it collects the whole 500 MiB working_dir and uploads it to the GCS. It can take 30s to do so - during which the Dashboard event loop is blocking.
This PR moves the uploading to another thread. This avoids event loop blocking.
This PR also removes a dead reference to gcs_client in http_server_head.py.