Has the sizing of the NSX-T managers changed in 3.2.x?
A colleague of mine recently completed a dual site NSX-T deployment that required the addition of two vCenters to the single NSX-T manager cluster, in his case the VC’s were running one at each site however the same issue would occur for a single site deployment the only thing that triggers the issue is adding more than one compute manager to a Medium sized NSX-T manager.
So whats the issue?
As you can see in the screenshot below there is a Capacity alert that is triggered when more than one Compute Manager is added to NSX-T. As you can see the Current inventory = 2 so its 200% over its allowed value as the Maximum capacity is 1.
If we take a look at the config maximums website
We can see that a Medium NSX Manager has support for 2 Compute managers
Interestingly this has been added to the 3.2.1 release notes as the 3.2.0 release notes only ever showed the maximums for Large managers with no detail on the other sizes however it has always been 2 Compute managers for Medium NSX-T Managers
So is there an issue or not?
Given that the config max conflicts with the Alarm received in NSX-T you can be forgiven for being confused as to whether there is an issue here or not. I recently read a blog post that after discussion with VMware GSS stated that there was an issue and that only 1 Compute manager was supported and as such the Managers would need to be upscaled to resolve. This didn’t sit right with me so I asked directly on the internal VMware systems. It turns out this is a know bug and that 2 compute managers are indeed fully supported. So for now either adjust the alarm definition to increase the capacity, disable the alarm or simply ignore it an carry on 🙂