Note: I’m not an admin, but I’m filling in for our out-sick admin…
We’ve had an interesting incident where failover occurred, even though no clusters have been defined in our system. HTTP failed on one of our servers, and it was brought down (i.e., the OS was rebooted). During that time when people tried to access their databases from the client on the downed server, rather than receiving a message stating that the server was down, new replicas were added their workspace for replica dBs on another server in the domain. I tested this with my development server, and was unable to duplicate it. Can anyone out there shed a little light on what might be going on? Is this standard functionality, or an anomaly?
Thanks,
- Yaron