SRM - Test & Recovery IP addressing

[NPBrown]

Hi, I'd like to know if it's possible to define different IP addressing schemes within SRM (6.1) depending on whether the activity is a test (on an isolated network) or actual recovery (onto the active/routed DR network). It doesn't seem like this is possible natively - if so then what's the best way to achieve this?

Thanks in advance.

[vbrowncoat]

SRM doesn't have that functionality built into it. If you wanted to do something like this you could use a script to change the address only during tests.

Why are you wanting to do this? The idea of a test is for it to duplicate the actual recovery plan as much as possible. Just curious.

[NPBrown]

I have an active, routed vlan that hosts our active DR systems (AD DCs, file servers, Citrix provisioning). During an actual failover our systems would be brought up on this vlan using the mapped IP addresses that are valid for this vlan. However, when I am testing I want these systems to be brought up on our isolated, non-routed, vlan on a different set of mapped addresses. In this instance I plan to have a XenApp desktop that is connected to both the active and isolated vlans, so that the desktop can be accessed by testers, but can also access systems inside the isolated vlan (****but does not route between them***). For this reason the systems on isolated network cannot use the "active" IP addresses on the isolated vlan.

I was assuming that I would have to script something. I'm considering putting together a script that uses the "VMware_RecoveryMode" environment variable to determine whether to run dr-ip-customizer.exe (i.e. in test mode). Does that sound the best way to go?

[vbrowncoat]

In my opinion the best way to go would be to have the test network completely isolated from you production networks (connected by a dual homed XenApp machine is fine) so that when you run a test you change the addresses to the same thing that you would when you run an actual failover. Doing this you should be able to have the same addresses in your isolated environment as in your production.

I would recommend doing whatever you can to make your test as like a failover as possible. This should increase your confidence in the recovery plan functioning as expected during an actual planned migration or failover.

[NPBrown]

Except that the dual homed XenApp desktop has to be able to communicate with systems on both the active and isolated networks. For that I need a different IP addressing scheme on each network - I cannot use the same IP addressing scheme on the isolated network as on the active network. That is why I need different IP address mappings for test and failover.

[vbrowncoat]

I would recommend figuring out how to put the XenApp desktop on a different internal network rather than using different IP addresses for test vs failover.