While a vast majority of systems come back online (from my experience 95%) some do not as the network stack was broken during the installation. After further research it appears QoS can be a major factor with this and the recommended work around is to disable QoS if it's not being used.
However I've been in multiple environments where QoS is being used. I've tested a script that disables QoS, installs HIPS, enables QoS but this caused 3 drops in network connectivity and required perfect syncing between steps. It was causing more headaches than it solved. Handling a couple systems with hands on remediation is straight forward, but it becomes very time consuming deploying to a environment of 30,000+ and ultimately requires someone to physically touch a system.
I'm curious to know how others deal with this issue. Hands on remediation? A better diagnosis to determine which systems are most likely to fail?
Thank you and forgive me if this is being discussed elsewhere.
Message was edited by: jkbharrison on 1/21/13 5:56:55 PM CST