Skip to content

SOLR-17754 Fix race condition in overseer main loop. #3350

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: main
Choose a base branch
from

Conversation

psalagnac
Copy link
Contributor

https://issues.apache.org/jira/browse/SOLR-17754

Description

Stuck overseer that sometimes happens under high load, when the overseer has at least 100 running tasks.

See Jira for the full scenario as description is pretty long.

Solution

This fixes the overseer main loop so we never submit more than 100 concurrent tasks to the thread pool. Instead of manually tracking when a task is complete, we check the status using a standard java Future.

The changes also makes sure we don't write the result to ZK response node when we should not (see comment), removing the erroneous occurrences of log "Response ZK path: <node> doesn't exist. Requestor may have disconnected from ZooKeeper"

Tests

Add a new test to make sure we don't fail anymore with lot of tasks.

This fixes the overseer main loop so we never submit more than 100 concurrent tasks to the thread pool. Instead of manually tracking when a task is complete, we check the status using a standard Future.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant