I'm scheduled to run a planned recovery on our Production servers this coming weekend. We will do a recovery, testing and then reprotect and fail back.
I've been through this with a group of development and test servers and it has gone well. The one question I have is about the site to site resource mappings, such as folder and network.
I have these defined from the protected site to the recovery site. In my testing with development servers it doesn't appear that creating mappings are needed from the recovery site to the protected site when doing a failback.
My assumption is that these are not needed as long as they are defined the other way. Is this correct?
Any other pointers about doing a test recovery are also appreciated.
You must configure inventory mappings on both sites before running reprotect, or reprotect might fail.
Take a look: How Site Recovery Manager Performs Reprotect
You must configure inventory mappings on both sites before running reprotect, or reprotect might fail.
Take a look: How Site Recovery Manager Performs Reprotect
That's interesting, in doing recovery with test servers I didn't have the resources mapped both ways and never had an issue.
I mapped them in the reverse direction before our planned recovery. I didn't have and problem running reprotect.
Now if only all the DNS entries would have updated dynamically as hoped for. :smileyangry:
That's interesting, in doing recovery with test servers I didn't have the resources mapped both ways and never had an issue.
I mapped them in the reverse direction before our planned recovery. I didn't have and problem running reprotect.
The reverse mapping is needed if there are any problems on the protected site that affect the original VMs. As part of the reprotect process the original VM compute and network locations are used by default. If however they can't be determined for whatever reason the reverse mappings are used instead.
Now if only all the DNS entries would have updated dynamically as hoped for. :smileyangry:
How are you planning to do DNS updates? For many customers the use of dynamic DNS alleviates this issue considerably. Here are a couple of links for how to do this with windows: Allow Dynamic Updates (2K8), How to configure DNS dynamic updates (2K3). If you can't use dynamic DNS there are some other options you can do with SRM involving script call-outs.
http://support.microsoft.com/en-us/kb/2520155
We do allow dynamic DNS updates but about 10% failed to register after the recovery and 20% or more failed on the way back.
There might be a Microsoft DNS client hotfix for this issue but it was published in 2011. So I would have think this would have been fixed by now!
I have no idea why the editor insisted on pasting the link at the top of the message!