Forum Discussion
Deb_Allen_18
May 31, 2006Historic F5 Account
Hi Andrea -
Hopefully you've solved the original persistence problem by now.
Regardless, you're actually pretty close on this, but no rule is required.
You will need to create an application health monitor to verify the health of each node: Click here
If you simply create a single pool with both members at differing priorities, with min_active_members set to 1, you should see only traffic to the higher priority member unless it has failed.
You can put more members in the pool at different priorities for additional fallback targets in case of multiple failures.
Since your app is stateful, you may still want to set persistence to prevent healthy sessions on node 2 from breaking state by jumping back to node 1 when it recovers.
Good luck!
/deb