HL7 TCP eWay Resource Usage

We have a number of HL7 eWays connecting to external systems. If one of these systems goes down and a message is sent through to the collaboration handling the eWay, the entire logicalhost seems to grind to a halt until the connection is remade.

Has anyone else seen this happen? In the future we may have up to 20 different clients with anywhere between 2 and 10 tcp connections, and for one to go down and kill the whole system is quite disconcerting.

Cheers,

Andrew.

[494 byte] By [AndrewShaw] at [2007-11-26 10:42:25]