We have configured two RADIUS servers using MySQL Replication to keep them in sync. Everything is working on the backend but I noticed the following behavior:
- When logging in to an AP, the request goes to the first RADIUS server. I expect this.
- If I take the primary RADIUS server down, the request goes to the 2nd server. I expect this.
- If I bring the primary server back up and take the secondary down, the request does not come thru the primary server.
I’m trying to understand the behavior of this. Does the AP mark the RADIUS server as unreachable for X amount of time then try it again after the time goes up? This is a partially live system so I can’t leave the RADIUS servers offline for more than a few minutes at a time.