Bottleneck recovery problem.

[john.cunningham]

                   We have had two serious incidents that have followed this pattern. An upstream link or device becomes saturated or unable to pass much traffic. The problem is cleared. The Ironport continues to have difficulty passing traffic with multi second response times. A reboot immediately clears the issue (we have 2 S360's, 2 wccp services with dynamic routing). Are there some tuning parameters we can use to prevent this? Why would only one of a load balanced set behave this way. Configurations are matched. Which logs would have the best chance of giving me a clue? Thanks for anything. jc

[Erik Kaiser]

Hi John,

If you have not already opened a case with the Content Security Team I would request that you do so. This particular issue is far to complex to solve on the forums.

Sincerely,

Erik Kaiser

WSA Forums Moderator