DNS_REQUEST

Your GTM should be able to load balance across your two data centers without an iRule. Here's an example: You have an FQDN, www.example.com, that you'd like to resolve to a server in data center A if it's available. If it's not available, you will instead resolve to a server at data center B. Your FQDN will be configured as a wide IP. To that wide IP, you'll assign a pool. Your pool should contain the two server IP addresses, one from data center A and one from data center B. The monitor you assign to the pool will need to be built to best reflect the availability of your two servers. The load balancing method that you setup on your pool should be type 'global availability'. Ensure your data center A server is first in your pool member order. This will ensure that it gets handed out as an answer for all queries if it's available. If it's unavailable, then the GTM will return the next available pool member, which will be data center server B. Hope I'm not rambling too much.