Failover between sites with Primary and Secondary NSX Manager Configuration - What about the controllers?

[niksheridan]

Hi forum,

Its still not clear to me, despite the documentation I have read what happens to the NSX controllers in the event the primary site goes completely offline.  I have a secondary vcentre in a linked mode, a primary and secondary NS manager, but my NSX controller cluster only exists at my primary site?

During a complete failover, do I need to manually create a new universal NSX controller cluster at the DR site?  Could this be automated with SRM?  Or most importantly whats the recommended approach?

Thanks

Nik

[Richard__R]

‌It's a manual redeployment into the recovery site after promoting the Manager there to a Primary. You then update the Controller state from the Manager and hosts will be notified that they need to talk to the new Controller Cluster. You can't protect the Controller Nodes with SRM.

[RaymundoEC]

‌I belive you can use an API call to automate the deployment after a failure and then put it in a vRO WF to invoke when disaster occurs.

[Richard__R]

‌Good point - and that's no doubt the best approach. When I say manual I mean you have to put something in to handle that as NSX won't natively orchestrate it.

[GyroDom]

Many thanks Richard,

I was about to ask this question too.

The only place it's listed in any documentation is on page 26 of 'Disaster recovery with NSX and SRM'

[NealeC]

Of course not! 🙂

Then VMware couldn't sell more VRO/VCO/VRA licenses 🙂

I think that the monitoring/alerting in NSX is still a bit "green" too.

As they make it more vmware and less Nicera I think that will improve though.

6.2.2. is a massive leap in bugfixes and traceflow is awesome!