Improve on_connected callback to avoid reconnect storm #292
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
Jason-Jian
approved these changes
Jun 14, 2019
anson627
force-pushed
the
anson-remove-connected
branch
from
0c929a4 to
ee71aab
Compare
allenlsy
approved these changes
Jun 14, 2019
anson627
changed the title
Ramyak
reviewed
Jun 14, 2019
anson627
force-pushed
the
anson-remove-connected
branch
from
0ada8a0 to
84da546
Compare
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Outdated suggestions cannot be applied.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.
The on_connected callback was originally introduced to handle an issue that zookeeper session got stuck (will not receive zookeeper watch event forever), during a connection lost and reconnect.
But we discovered some side-effect of this fix, there are multiple on_connected callback get triggered for a single disconnect and reconnect (here seems other people are facing the same issue: zk-ruby/zk-eventmachine#2). And since we do reset zookeeper watcher by read everything. This creates substantial load on zookeeper.
Here are two optimizations introduced by this change:
@austin-zhu @Jason-Jian @allenlsy @Ramyak
Before (reboot zookeeper twice, got 3 on_connected each time):
After (reboot zookeeper server once, got watcher refresh once):
reboot zookeeper server twice, only got watcher refresh twice
