This morning we noticed that one of our AMDs was getting a large number of network errors, however, NAM (DCRUM 2017) didn't initiate a failover. I couldn't find a way to manually fail it over, so I went into the CAS and removed the one with the errors from the data stream. So my questions:
1. Is there a way to do a proper failover manually in such cases?
2. What is the criteria for an automated failover?
Solved! Go to Solution.
Answer 1: In fact the change between feeding NAM server from main probe and failover probe can not be performed manually (unless you change configuration the way you described above), it is triggered after meeting necessary conditions, which leads us to ...
Answer 2: NAM server will send requests to failover probe either when it can't connect to main probe, or when it receives codes different than 200-299 or 401 as a response to its request.