Skip to content

Remove a potential deadlock when shutting down a RoutingTableProvider.#1751

Merged
jiajunwang merged 1 commit intoapache:masterfrom
jiajunwang:master
May 26, 2021
Merged

Remove a potential deadlock when shutting down a RoutingTableProvider.#1751
jiajunwang merged 1 commit intoapache:masterfrom
jiajunwang:master

Conversation

@jiajunwang
Copy link
Contributor

@jiajunwang jiajunwang commented May 21, 2021

Issues

  • My PR addresses the following Helix issues and references them in the PR description:

Fixes #1748

Description

  • Here are some details about my PR, including screenshots of any UI changes:

This PR aims to avoid a potential deadlock that happens in a race condition when shutting down RoutingTableProvider.

The race condition happens when the RoutingTableProvider is shut down and the corresponding HelixManager is processing an onLiveInstanceChange event. Both threads require 2 locks, HelixManager lock, and RoutingTableProvider._lastSeenSessions, in different orders.
To resolve this problem, this PR changes 2 logic.

  1. Remove the _lastSeenSessions lock usage. So no more deadlock. The logic correctness is now ensured by the AtomicReference.getAndSet method.
  2. Remove LiveInstanceChangeListener in the shutdown call before processing the other logic. This is a missing logic and should be done regardless of the deadlock problem. Moreover, this change also avoids a more complicated race condition introduced by the previous change that may lead to CurrentStates callback handler leakage.

Test cases in TestZkCallbackHandlerLeak have been updated to validate the logic changes.

Tests

  • The following tests are written for this issue:

TestZkCallbackHandlerLeak

I run the test over the weekend, passed continuously 3533 times already.

  • The following is the result of the "mvn test" command on the appropriate module:

(If CI test fails due to known issue, please specify the issue and test PR locally. Then copy & paste the result of "mvn test" to here.)

Changes that Break Backward Compatibility (Optional)

  • My PR contains changes that break backward compatibility or previous assumptions for certain methods or API. They include:

(Consider including all behavior changes for public methods or API. Also include these changes in merge description so that other developers are aware of these changes. This allows them to make relevant code changes in feature branches accounting for the new method/API behavior.)

Documentation (Optional)

  • In case of new functionality, my PR adds documentation in the following wiki page:

(Link the GitHub wiki you added)

Commits

  • My commits all reference appropriate Apache Helix GitHub issues in their subject lines. In addition, my commits follow the guidelines from "How to write a good git commit message":
    1. Subject is separated from body by a blank line
    2. Subject is limited to 50 characters (not including Jira issue reference)
    3. Subject does not end with a period
    4. Subject uses the imperative mood ("add", not "adding")
    5. Body wraps at 72 characters
    6. Body explains "what" and "why", not "how"

Code Quality

  • My diff has been formatted using helix-style.xml
    (helix-style-intellij.xml if IntelliJ IDE is used)

This PR aims to avoid a potential deadlock that happens in a race condition when shutting down RoutingTableProvider.

The race condition happens when the RoutingTableProvider is shutdown and the corresponding HelixManager is processing an onLiveInstanceChange event. Both threads require 2 locks, HelixManager lock and RoutingTableProvider._lastSeenSessions, in different orders.
To resolve this problem, this PR changes 2 logics.
1. Remove the _lastSeenSessions lock usage. So no more deadlock. The logic correctness is now ensured by the AtomicReference.getAndSet method.
2. Remove LiveInstanceChangeListener in the shutdown call before processing the other logics. This is a missing logic and should be done regardless of the deadlock problem.Moreover, this change also avoids a more complicated race condition introduced by the previous change that may lead to CurrentStates callback handler leakage.

Test cases in TestZkCallbackHandlerLeak has been updated to validate the logic changes.
Copy link
Contributor

@junkaixue junkaixue left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Overall, looks good

@jiajunwang
Copy link
Contributor Author

Approved by @junkaixue , I will merge the PR.

@jiajunwang jiajunwang merged commit faa0cd3 into apache:master May 26, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Deadlock in RoutingTableProvider shutdown process

2 participants