Running VRA version 7.4.0 (build 8182598), NSX is version 6.4.1 Build 8599035. I really haven't been able to identify a pattern but perhaps 1 out 5 items, or blueprints that have been requested from the catalog and have been in use for days/weeks will fail with that error message when the Destroy request is sent.
The work around is tedious, you can "force destroy" the item, however there are several artifacts left behind in vSpehe and NSX, primarily networks and edge-routers. So you have to map out what's connected and delete in the correct order, a real PITA. This is happening with blueprints that use NAT networks, haven't tested other types of networks since the NAT type is what we're currently using.
Any insights would be appreciated.
Alex
The same issue happens for some deployments using the vRA REST API, sadly.
This problem was solved in 7.4 HF5 and 7.5
NSX deployed items periodically show up as virtual machines and not NSX items. | NSX deployed items periodically show up as virtual machines and not NSX items | HF5 | 7.5 |
It happens when the inventory collection is run at the same time as an Edge deployment