Let me get this straight.. you have 2 different F5 instances, and there's one that's configured to forward traffic to VIPs exposed by the second one, correct?
There's 4 different HTTPS pools on different ports, all pointing to a VS on the second unit, which runs them on a single IP with different ports.
You're using the same HTTPS (standard? custom?) monitor on all pools, but two fail and two don't.
First thing that comes to my mind: on the "back end" F5, do all VS/pools forward traffic to the same nodes?
If you're using a custom string, you should also make sure that your monitor points to a resource that exists on all 4 services, ex. if you run "GET /health" then /health page should exists on all 4 services.
Also, if back end server are not the same, you should tune the "Host:" header in your monitor, in order for traffic to match.
The easiest way to test your monitor is running a curl command or something similar, via F5 cli :
(echo -e "GET /health HTTP/1.1\r\nHost: www.f5.com\r\nConnection: Close\r\n\r\n";sleep 1) | openssl s_client -connect 10.0.0.10:443
Hope this helps
CA