GTM Setup with Dual Data Centre
I wonder if someone could throw a couple of config ideas my way. I'm familiar with both LTM and GTM but less so with the GTM and having worked my way through the config guide I still didn't get an answer to my issue.
I have two F5 GTM devices each located at a different data centre. These two data centres are connected by L2/L3 links. VLANS/IP subnets from one data centre also exist at the other data centre. The datacentres host VMWARE servers where a server can be moved and made functional at either data centre. I have an internet pipe at each data centre which is provided by different vendors so IP addressing is unique for the internet pipe at each DC.
My issue is (sorry long winded):-
The GTM's are set up to send out two IP addresses for any WIDE IP that I have configured, one IP is a public IP address for DC1 e.g. 22.214.171.124 and one ip address a public address for DC2 e.g. 126.96.36.199. Internally both these addresses belong to the same VM server e.g. 10.1.1.1.
I want to health monitor this server and also the internet pipe so I have a health monitor setup that makes sure 10.1.1.1 is up and a health monitor set up at each site to ping the internet gateway for that site e.g. 188.8.131.52 at DC1 and 184.108.40.206 at DC2.
I notice that the health monitor only tests from the GTM that is associated with the datacentre that the server belongs to which sounds logical. So if 10.1.1.1 belongs to DC1 then the GTM at DC1 send out the health monitor. However in my case the GTM at DC2 can also reach 10.1.1.1. So what I would like is for GTM1 to offer out 220.127.116.11 or 18.104.22.168 addresses if its internet pipe is up and it can see 10.1.1.1. GTM2 to offer out 22.214.171.124 or 126.96.36.199 addresses if its internet pipe is up and it can see 10.1.1.1. If the internt pipe at say DC1 fails then 188.8.131.52 is to be withdrawn. If access from DC1 to 10.1.1.1 goes down then again withdraw 184.108.40.206 but things should still exist from DC2.
What I want to know is whether there is a best practise or standard way to setting up a health monitor for this scenario where we have the same server and address at the two data centre.