We’re seeing a flood of CLOSE_WAIT connections exceeding the Axigen limits and bringing the client interfaces (mail ports and webmail) down blocking further connections by clients. Admin continues to work as usual so its a software lock rather than global firewall block. Initially we thought it was simply a resource issue and we increased connections to allow for the increase in use, but this was quickly exceeded also, then we realised the CLOSE_WAIT filling netstat, so the sessions arent clearing on the server. Is there a fundamental issue with the app not closing TCP sessions or is this a symptom of some other issue? I couldn’t see any lag from SAN (SSDs) that might be co-morbid with a cascading app close failure, io time was low, in fact Axigen has been superlatively well behaved in terms of resource usage, it barely puts a dent in the system with a Load Average of around 0.1-0.3 .
Anyway hope you gents can advise, we’re still auditing it but thought you might be able to provide some insight 
thanks!
We’re running the update just before the most recent update.
Brad.