In version 1.1-4, the call to "latencies.remove()" (LatencyAwareHClientPool line 48) may throw a NoSuchElementException if the "resetThread" (See DynamicLoadBalancingPolicy instantiation line 62-74) clear the "latencies" LinkedBlockingDeque just before.
Using the poll() method instead of remove() should fix the issue.
How can i submit a patch ?
Looks like it's been fixed.
Any chance this will be ported into a 1.1 release? Or is support for Cassandra 1.1 essentially deprecated?