I'm attempting to develop an NSX-v load balancing configuration with the following requirements:
1. Web1 will be primary web server and should be used if up
2. If Web1 is down, use Web2
3. If Web1 and Web2 are down, redirect to sorry server URL.
This essentially and active/passive configuration with a sorry server as the tertiary. This seems straight-forward in NSX-T, but not so much with NSX-v.
I have been working with application rules and have gotten 1 & 2 to work together, and 1 & 3 to work together, but I'm struggling to find the appropriate syntax to get all 3 criteria to work together. Any insight would be helpful.
Answer was in plain sight. Set the weight value on the primary pool member to a value higher than that of the second pool member. In this case, set Web1 to 100 and Web2 to 1. Then, leverage an application rule to do a redirect to the sorry server when the pool has-0 members available