When an actor is registered, a quasar actor mbean is created, this mbean remains even after the actor has terminated. On systems that register a large number of actors, this can result in a memory leak.
My current workaround is to use LocalActor/stopMonitor when deregistering or terminating an actor.
Is there a better practice or, failing that, are there API improvements we can make to clean up reliably?
When an actor is registered, a quasar actor mbean is created, this mbean remains even after the actor has terminated. On systems that register a large number of actors, this can result in a memory leak.
My current workaround is to use
LocalActor/stopMonitor
when deregistering or terminating an actor.Is there a better practice or, failing that, are there API improvements we can make to clean up reliably?