In my case I upgraded from 3.0 to 188.8.131.52 only using the MUB and uploaded it. I use the procedure to upgrade from 2.5 to 3.0 and everything worked because is all the same. Maybe the console changed a little bit but all the same options are there.
How are you?
But on your upgrade did you add the 100GB disk on all Manager appliances?
I did not because I already had the secondary disk with 100GB so I guess this is an actual requirement when you come from NSX-T 2.x
Already done the update from 3.0.1 to 3.0.2. Everything ok.
Wasn't necessary to add the 100GB disk to each manager. Just upload the .MUB file, run the pre-checks (just a warning about the missing backups). In my case all the edge nodes and managers share the same vsphere cluster, so, what I needed to grant before, was the DRS in fully automated mode. The upgrade process done everything (migrate vm's and put the hosts in maintenance mode).
On the last step of the update (managers update), at the end we will lost the communication. It is normal because the managers will be rebooted so we will need to wait several minutes until we can reload the browser and get back to the page again.