Custom HTTP monitor questions

When you create a HTTP monitor for a basic website. What should you look for send/get,Whats the best common practice? Just looking for a word in the html page?

The best practice is really dependent on the application. Generally speaking, the more specific to the application the value in the HTTP response, the more robust the monitor. You can look for a 200 response status code, or a standard HTML attribute tag, but that's pretty generic. Or you could look for some text value that 1) always exists in the application responses, but only if 2) the application is fully functional.

When you do send string for HTTP monitor, how can you see the response for debugging and to determine what would be options for receive strings? Do you just telnet to the port or does f5 have something for debugging

If you're not re-encrypting to the server, you can use a standard tcpdump on the interface/VLAN that traffic leaves to reach the server:

tcpdump -lnni [interface, VLAN, or 0.0] -Xs0 [and any additional filters]

The -Xs0 parameter will dump the payload to the capture output.

1. If you have a website that contacts a server on a certain port, how would you do a monitor for that website For example www.aaa.com run a yyy.jsp file on port XXXX to a backend server How would a write a http monitor on aaa.com to check if that file is working

Interesting. One idea would be to monitor some content of that yyy.jsp file from www.aaa.com. Another idea might be to create a separate "test" page in your application that checks things like application health, backend server file access, etc. and then reports back up/down status to the monitor.