I understand why the default configuration of nettcp binding to a WCF service through a load balancer is not a great idea. The bottom of this msdn page has the explanation why.
But the last line of that section says "For best performance in load-balanced scenarios, consider using NetTcpSecurity (either Transport or TransportWithMessageCredential)." The link to the page about NetTcpSecurity says nothing about load balancing at all.
I am trying to understand if the documentation is implying that by using transport level security it somehow negates the need to alter the connection pool timeouts. Or if it is merely suggesting that adding transport level security to your connection while also altering the timeout configuration would somehow be more performant than just altering the the timeout (which would be weird, how would adding overhead make it more performant?)
Does the SSL tunnel force connection teardown after each request or something? That's the only thing I can come up with as to why this could possibly remedy the problem.
We currently have two different products running nettcp over ssl out the back side of the proxies and through a load balancer to our internal services and haven't noticed anything out of the ordinary. Also our services run sessionless and stateless.
