Hi Kevin, I think re-creating the Set is much faster as well. I've looked at the implementation of org.apache.commons.collections.map.AbstractHashedMap.clear() and it simply iterates over all entries in the hashmap and sets all references to null. Btw, as you can see in the backtrace (AbstractHashedMap_clear()run2.png) each clear is being called in BrokerImpl.endOperation(), so this would be a great performance improvement in the OpenJPA code.
kind regards, Christiaan -- View this message in context: http://www.nabble.com/Performance-drop-in-AbstractHashedMap.clear%28%29-tf4769771.html#a13656730 Sent from the OpenJPA Developers mailing list archive at Nabble.com.
