At the weekend our F5 load balanced CAS array name was moved to our 3rd datacentre and all users lost access to Outlook. The databases were running on DAG2 listed below at the time.
We have 3 DAG servers configured as the following:
DAG SERVER ONE (DAG1) | Datacentre 1 | Primary AD Site | Multi role server MBX, CAS, HT
DAG SERVER TWO (DAG2) | Datacentre 2 | Primary AD Site | Multi role server MBX, CAS, HT
DAG SERVER THREE (DAG3) | Datacentre 3 | Secondary AD Site | Multi role server MBX, CAS, HT
We primarlily run from DAG1 or DAG2 - DAG3 is for real DR and loss of Datacentre 1 and 2. The CAS array "Outlook" is configured for the primary AD site load balanced across DAG1 and DAG2.
We use an F5 load balancer that balances the connections across DAG1 and DAG2, if both fail then DAG3 is used automatically. We have tested and it works fine simulating loosing DAG1 and DAG2 and bringing up as a 1 node cluster on DAG3 with a FSW and the load balancer pointing at DAG3 for Outlook connections.
For some reason this stopped Outlook users connecting - can anyone advise me why as I have been asked to test this again and I was expecting it to work?