Improve recovery #245
Merged
Improve recovery #245
Conversation
This file contains hidden or 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
Create a locator connection for each provided URI. This way a connection can take over when the locator node goes down. This speeds up recovery. Track scheduled tasks. This is likely to be disabled in a stable release. Useful to track down unfinished tasks. Add retry to operations in the consumer coordinator. Refresh consumer candidate nodes if the re-assignment of a consumer times out. This improvements are based on the feedback from the effects of a rolling restart in K8S using stream-perf-test. Not all producers and consumers are recovered after all nodes have been restarted. The changes in this commit mitigates this problem.
Another recovery can kick in while a first one is in progress. As we added retry, the first one can retry and refresh its connection, so both will succeed and we end up with more consumers than expected.
acogoluegnes
force-pushed
the
recovery-improvements
branch
from
065fd49 to
10a0887
Compare
When stream is not available and even when there is no candidates at the moment (added some retry delay for the latter).
It would contain the client connections for a given node. It was difficult to maintain the consistency between those structures, so we're better off not using this layer. Now client connections are looked up with a linear scan, which is good enough, as the number of connections should remain under the thousand.
It would contain the client connections for a given node. It was difficult to maintain the consistency between those structures, so we're better off not using this layer. Now client connections are looked up with a linear scan, which is good enough, as the number of connections should remain under the thousand.
Instead of wrapping them in StreamExceptions. Easier to deal with timeout exception.
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.
Create a locator connection for each provided URI. This way a connection can take over when the locator node goes down. This speeds up recovery.
Track scheduled tasks. This is likely to be disabled in a stable release. Useful to track down unfinished tasks.
Add retry to operations in the consumer coordinator.
Refresh consumer candidate nodes if the re-assignment of a consumer times out.
This improvements are based on the feedback from the effects of a rolling restart in K8S using stream-perf-test. Not all producers and consumers are recovered after all nodes have been restarted. The changes in this commit mitigates this problem.