As a vSAN administrator, I recently encountered an issue during a deployment of VCF 5.0. After the deployment failed, I was left with a vSAN datastore on the first host in the cluster, which blocked me from retrying the deployment. The greyed out option to delete the datastore made it difficult to proceed further.

To resolve this issue, I had to SSH into the host and get the cluster. First, I needed to obtain the Sub-Cluster Master UUID, which I copied to the clipboard. To leave the cluster, I used the command:

esxcli vsan cluster leave -U

This command successfully left the vSAN cluster and I was able to delete the datastore and its partitions on the disks. The ESXi UI now showed that the datastore was gone, and when I checked the CLI, there were no longer any vSAN clusters or storage configured on the host.

After successfully deleting the datastore and leaving the cluster, I was able to retry the deployment of VCF 5.0. This time, the deployment was successful, and the VCF management cluster was deployed without any issues.

This experience taught me an important lesson about the importance of carefully planning and executing vSAN deployments. It also highlighted the need for a thorough understanding of vSAN concepts and commands to troubleshoot and resolve issues that may arise during deployment.

In conclusion, if you are facing a similar issue with a failed vSAN deployment and a stuck datastore, I recommend attempting to leave the cluster using the command esxcli vsan cluster leave -U . This should allow you to delete the datastore and its partitions, and then retry the deployment of VCF 5.0. Remember to carefully plan and execute your vSAN deployments to avoid any potential issues or failures.