Uploaded image for project: 'OpenShift Request For Enhancement'
  1. OpenShift Request For Enhancement
  2. RFE-3464

Feature to wipe disks during ZTP cluster's provisioning and deprovisioning

XMLWordPrintable

    • False
    • None
    • False
    • Not Selected
    • 0
    • 0% 0%

      1. Proposed title of this feature request

      Feature to wipe disks during SNO cluster's Installation

      2. What is the nature and description of the request?

      Since we're for a production environment need the Zero Touch Provisioning process to be fully automated, no manual steps can be involved for cleaning the disks and the process should ideally be driven from the ZTP SiteConfig

      Features:
       - wipe selected disks during installation, and when de provisioning servers.

      • Need the ability to preserve partitions during installation.

       - wipe all the disks during installation, and when de provisioning servers. 

      Because, there would be different server's configurations on a same clusters, the Siteconfig should act at Node Level. Something like:

            - hostName: "master-2.org"
              role: "master"
              rootDeviceHints:
                deviceName: "/dev/nvme0n1"
              wipeDisks:
                - /dev/nvme0n1
                - /dev/nvme0n2
                - /dev/sdc  
           - hostName: "master-3.org"         
             role: "master"         
             rootDeviceHints:           
               deviceName: "/dev/nvme0n1" 
             wipeDisks: 
               - /dev/nvme0n1 
               - /dev/nvme0n2
           - hostName: "worker-1.org"         
             role: "worker"         
             rootDeviceHints:           
               deviceName: "/dev/nvme0n1" 
             wipeDisks: [*]
        
      

       

      3. Why does the customer need this? (List the business requirements here)

      After the cluster is installed, they need to have cluster ready for installing ODF.  If any stale data is discovered (maybe from a previous cluster installation) on the OSD disks, the ODF process will detect this and fail.
      They are working on fully Zero Touch Provisioning. Now, they have to manually connect to servers to wipe the disk, before continuing with their ODF Installation.
      It seems an interesting feature to choose to wiped disks during the installation and during de-provisioning.
      Apart from this ODF concrete case, it seems interesting to have the ability to provision/deprovision clusters on a clean state. So any, (re)installation will act always the same, no matter what was installed there previously.

      4. List any affected packages or components.
      ZTP, Assisted Installer.

            bweidenb@redhat.com Bradd Weidenbenner
            jgato@redhat.com Jose Gato Luis
            Votes:
            0 Vote for this issue
            Watchers:
            16 Start watching this issue

              Created:
              Updated:
              Resolved: