Looking at the stack traces I see that you are using bulk get request with a limit. Looking more closely at the code it seems that when a limit is applied the iterator is not properly closed and if this cache is distributed it will cause a distributed iterator to be left around which can cause a thread to be stuck. I have logged [ISPN-6923] HotRod bulk commands that take a count need to close the iterator - JBoss Issue Tracker to fix this.
By the way most of the threads here are fine, they are just in a BLOCKED state waiting for a task. It looks like the above bug though caused the async thread pool to fill up.
In my application i am using method putAllAsync(Map map) --> i guess you meant this operation.
Anyhow, is there a workaround for this issue (i.e enlarge some thread pool size, etc...)? Should i continue using bulk operation like putAll? Or shell i use regular put until this bug is fixed?
How often this can be?