- You have NSX embedded installed with a 3-node NSX cluster in place.
- You have just renewed the certificates in use by vCenter Server
- The Compute Manager connection in the NSX UI shows status down as seen in the following screenshot:
- Upon editing the connection, and updating it, the connection status stays down
- The following loglines are seen in the cm-inventory.log file of the NSX Manager:
registration_error {
error_message {
value: "Compute Manager ###### registration failed because of Unknown error in processing request for CM ########.### with id ########-####-####-####-########."
}
VMware NSX
This is a result of the use of the NSX embedded with a 3-node cluster.
Convert your NSX embedded deployment into a regular 3-node NSX Manager cluster. This can be done by removing the existing NSX Managers and redeploying three new ones by restoring them from backup.
It maybe possible to deploy an alternate resolution with the help of Broadcom Support, if you already have a three node cluster deployed from a pre-existing NSX embedded manager.
If you open a case with Broadcom Support, please upload the following information to it:
- Log bundle from all 3 nodes of the NSX Manager cluster
- vCenter Server log bundle
- Screenshot of the Compute Manager UI tab