Details
-
Bug
-
Resolution: Not a bug
-
Major
-
None
-
1.0.7, 1.0.13
-
None
-
Websphere Application Server 8.5.0.1 Network Deployment
Description
We use AsyncAppenders in our WebSphere AppServer 8.5 ND environment.
When an application starts, that uses logback, threads are created with name AsyncAppender-Worker-Thread-XXX for every defined async appender, as they should.
But we noticed, that when the application is stopped, these threads do not dissapear. And when the application is started again, the previously created threads all stay there, and new ones are created, and old threads never die. With a couple of application restarts, it's easy to get to houndreds of running worker threads, in the end, exhausting the native heap and killing the server.
Originally we used logback 1.0.7, now we tried it with 1.0.13, but it's the same.