As a retired Network Geek I don't know how many times I've seen this very same problem over the twenty odd years I've trouble shoot networks. In the Mainframe era of SNA this was the most common problem. That is why we invented response time monitors and application capacity parameters. Then Wans and Lans came along and this old technology was discarded because slow reponse oould never happend again, "Right". Maybe it is time use this old technology again with formalized Problem Determination Processes, just like the old days? Why do we always Have to Reinvent Wheel OVER and OVER Again?

Fitz