Clean up/remove NSX manager entries from SDDC database
book
Article ID: 439927
calendar_today
Updated On:
Products
VMware SDDC Manager / VCF InstallerVMware NSX
Issue/Introduction
Failed NSX manager with single node (not a cluster) and no backups to recover from.
Cause
Needing to re-deploy NSX to VCF environment due to failed NSX node(s) and no way to recover from backups.
Resolution
Take a snapshot of the SDDC manager.
Log into UI portal for the NSX Manger that needs to be redeployed.
Un-prepare vCenter clusters from NSX
Navigate to "NSX Manager UI > System > Fabric > Hosts".
Under Clusters tab, select the clusters from the impacted vCenter.
Verify that all hosts are in a "Not Configured" NSX Configuration state before proceeding to next step.
Remove NSX Compute Manager.
Navigate to "NSX Manager UI > System > Fabric > Compute Manager".
Click on 3 dots on left side of the compute manager for the impacted vCenter. Click on Delete option.
Delete NSX VM(s) from the vCenter.
Power off the NSX Managers / NSX Edges
Navigate to "vCenter UI > NSX Manager VM > Actions > Power > Power Off" and Confirm Power Off.
Once the VMs are powered off, select the NSX Manager VM and delete it by navigating to "vCenter UI > NSX Manager VM > Actions > Delete from Disk" and Confirm Delete.
Remove the NSX domain from the SDDC manager inventory.
SSH into the SDDC Manager with vcf and then elevate to root with su.
Collect the domain ID for the domain that will be reimported.
curl -i -X GET http://localhost/inventory/extensions/vi/domainInventory
Remove the NSX Manager from the domain. (If NSX Manager is shared skip this step)
curl -i -X DELETE http://localhost/inventory/extensions/vi/nsxtclusterdomains/<domain-id from above output>