You are viewing a plain text version of this content. The canonical link for it is here.
Posted to notifications@zookeeper.apache.org by GitBox <gi...@apache.org> on 2019/12/16 16:30:29 UTC
[GitHub] [zookeeper] symat opened a new pull request #1184: ZOOKEEPER-3651:
try to fix flaky NettyServerCnxnFactoryTest
symat opened a new pull request #1184: ZOOKEEPER-3651: try to fix flaky NettyServerCnxnFactoryTest
URL: https://github.com/apache/zookeeper/pull/1184
The testOutstandingHandshakeLimit is flaky, I tried to fix it in this commit.
- I added extra comments and did some restructuring in the code.
- Avoiding to start unnecessary ZooKeeper servers for tests don't require it
- Decreasing the number of client connections the test tries to initiate
- Increasing the timeout to make sure the connections get established
- Filtering the 'SyncConnected' events in the client watcher to make sure
the given connection is really established before counting it
I think the last two points above should fix the flakiness. I tried to run the
test in docker, and before the fix it failed for me once in every 4-5 execution.
After applying these changes I re-executed it 100 times without failure.
If these fixes are not enough, then we can introduce some only-visible-by-test
method to add sleep in the SSLHandshake process in the production code to
force to have handshakes in parallel. However, it would be nice to avoid that.
Let's hope that these fixes will be enough.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
With regards,
Apache Git Services