Solved: ISE - Best Practices For Restoring After Network Crash

Quintin.Mayo · ‎10-14-2019

Hi,

Last week we had one of our 6807XL crash on Thursday 10/10. As we were troubleshooting the issue with the 6807XL, we noticed some issues with ISE not coming back up correctly after the crash. We want to review the best way to recover or diagnose issues going forward with ISE. Currently we have been rebooting both sides of our ISE cluster, if we experience any issues, but is this the best procedure? Any information would be greatly appreciated.

Thanks,

Damien Miller · ‎10-14-2019

It actually sounds like there is a different issue here that is related to ISE, but the problem shouldn't be with ISE itself. When a switch fails, radius should recover on its own without having to do anything on ISE.

Having to reboot ISE indicates to me that there could be communication or config issues impacting NAD failover.

So the best practice should be more related to the network devices, and assuming ISE was deployed in a HA way, then it should be a none issue. You don't need to reboot ISE nodes under normal conditions unless something in software/hardware fails with ISE itself.

View solution in original post

Damien Miller · ‎10-14-2019

It actually sounds like there is a different issue here that is related to ISE, but the problem shouldn't be with ISE itself. When a switch fails, radius should recover on its own without having to do anything on ISE.

Having to reboot ISE indicates to me that there could be communication or config issues impacting NAD failover.

So the best practice should be more related to the network devices, and assuming ISE was deployed in a HA way, then it should be a none issue. You don't need to reboot ISE nodes under normal conditions unless something in software/hardware fails with ISE itself.