Need help. HTTP monitor based on hostname

Hi,

When I create a basic HTTP monitor and use a hostname the result is that the actual ip address of the host is used in the monitor. A one time nslookup is done.

If the ip address of the host changes, the health monitor will fail.

Is there a way to use the hostename only all the time, by doing an nslookup every time?

I expect it can be done by an external script, but I was not able to find an example on DevCentral.

Please help!:smiley:

Does the web server have virtual host or host header definitions that require a specific hostname be inserted in all requests? I wasn’t aware that any name resolution would be done for a monitor send string. I would have guessed that the static string would be sent in the monitor request. Can you clarify what you’ve tested and which version you’ve tested on?

Thanks,

Aaron

I use this a health monitor to check the status of the ISP link on a Link Controller. I want to check that some internet web pages are reachable by just doing a “GET /”.

If I configure the basic http monitor I type for instance www.nu.nl as the adress to be checked. If the monitor is now created I see that the ip address is put in there instead of www.nu.nl.

}

monitor HTTP-Nu {

defaults from http

transparent

dest 62.69.179.231:http

}

Now if www.nu.nl changes the ip address, the monitor will still check the old address and therefore fails.

So I need a script that can do an dns lookup first and then do the check on the resolved ip address.

BIGIP is running version: 9.4.5 with HF2 (soon to be upgraded to 10.0.1

I think you might need to use an external monitor. It would be a fairly simple one to implement using curl as long as you have DNS resolution configured. curl accepts a hostname and will perform name resolution before sending the request.

The Monitoring Codeshare has several examples of external curl-based HTTP monitors.

Aaron