nadavFux opened a new pull request, #162:
URL: https://github.com/apache/flink-connector-elasticsearch/pull/162

   https://issues.apache.org/jira/browse/FLINK-40381
   
   In flink-connector-elasticsearch8, Elasticsearch8AsyncWriter.close() 
previously called this.esClient.shutdown(). In the Elastic Java API Client, 
shutdown() returns an ElasticsearchShutdownAsyncClient object for calling 
Elasticsearch Cluster Shutdown REST API endpoints (/_shutdown). It does NOT 
close the underlying RestClient or HTTP Async Client IO reactor thread pool 
(I/O dispatcher).
   
   As a result, every sink instance teardown on a Flink Session Cluster leaks 4 
zombie I/O dispatcher threads in the TaskManager JVM, accumulating until 
process thread limits are hit.
   
   Fix: Update Elasticsearch8AsyncWriter.close() to call 
esClient._transport().close(), which cleanly closes the underlying HTTP 
transport and terminates reactor worker threads.
   
   


-- 
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.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to