This repository has been archived by the owner on Feb 18, 2021. It is now read-only.
Upstream patches #9
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
Reviewers: samar Reviewed By: samar
Summary: Remove all references of UseWebsocket uConfig value, and related logic. Test Plan: Unit tests Reviewers: aravindv, gbailey Reviewed By: gbailey
Summary: use metadata-metrics client in output, input, store and replicator Reviewers: samar, aravindv Reviewed By: aravindv
Summary: This diff handles store failures for remote extent. Currently we only handle primary replica going down. Reviewers: kiranrg, gbailey Reviewed By: gbailey
Reviewers: nx Reviewed By: nx
Reviewers: gbailey Reviewed By: gbailey
Reviewers: aravindv, samar, gbailey Reviewed By: gbailey
Reviewers: samar, aravindv Reviewed By: aravindv
Summary: As title, this diff will disable soft retention for multi_zone extents in source zone Reviewers: kiranrg Reviewed By: kiranrg
Summary: While investigating the goroutine explosion, I noticed that putMessageBatch can potentially be stuck forever waiting for acks. We should ideally timeout using our usual timeout of a minute. Test Plan: make Reviewers: samar, venkat, weihan Reviewed By: weihan
Reviewers: samar, aravind, weihan Reviewed By: weihan
…on job after rpm is bootstrapped Reviewers: kiranrg, aravindv Reviewed By: aravindv
Summary: The stress test was successful and the latency was good as well after the change to use channel based close. This patch simply re-reverts all the leak fixes we already fixed. I will send the CR for the channel based close separately. This patch gets all the leak fixes by re-reverting the commit. One Unit test needed to be fixed to get the appropriate metrics reporter to make sure it doesn't crash. Test Plan: UTs Reviewers: weihan Reviewed By: weihan
Summary: 1. make sure we return immediately if timeout happens(previously it's possible we still read message from channel after timeout). 2. use a hardcoded large timeout for writing to msg cache Reviewers: aravindv Reviewed By: aravindv
Summary: This makes sure we don't have all callbacks competing for the path mutex. With the callback based close, there were a bunch of cbs competing with the path mutex and the latencies started getting worse. This patch just moves to a more idiomatic channel based close which makes sure we unload properly and also not make the latency worse. Test Plan: make Reviewers: weihan Reviewed By: weihan
|
|
datoug
approved these changes
Jan 10, 2017
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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.
Get all the relevant fixes for cherami-server and push it upstream.