Uploaded image for project: 'OpenShift Bugs'
  1. OpenShift Bugs
  2. OCPBUGS-31256

Baremetal IPI cluster failed to deploy

XMLWordPrintable

    • No
    • 1
    • Metal Platform 251
    • 1
    • False
    • Hide

      None

      Show
      None

      Description of problem:

      Facing an issue while deploying a baremetal IPI cluster [4.12] 
      
        -  Restarting the third master node aids with cluster construction if the first two control plane nodes manage to progress in less than 20 min; nonetheless, this behavior is inconsistent.
      
        -  However, most of the time only one node is managed to progress while the others did not. The installation eventually fails with a timeout error, and the ironic log displays a recurring cycle of multipath and ipmi errors.
      
        -  Attempted to turn off LLDP, however it had no effect on cluster deployment.

      Version-Release number of selected component (if applicable):

       

      How reproducible:

          Easily reproducible

      Steps to Reproduce:

          1. Baremetal IPI Installation. 
          
          

      Actual results:

          Installation failed with a repeating cycle of multipath errors and ipmi errors in ironic logs.

      Expected results:

          Installation gets completed successfully 

      Additional info:

      Infra details :
      - master nodes are all HP DL360 G10s
      - Firmware: ilo5 3.01.
      - They are using the Embedded ALOM cards which are HP Ethernet 10Gb 2-port 560FLR-SFP+ Adapters.  
      
      Log bundle and install-config file is uploaded [1]
      [1] https://drive.google.com/drive/folders/1K4F4NkCixz4Drhyw2WUO8o-1r_pIdrCj
      
      

              rpittau@redhat.com Riccardo Pittau
              rhn-support-rospatil Roshni Patil
              Jad Haj Yahya Jad Haj Yahya
              Roshni Patil
              Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

                Created:
                Updated:
                Resolved: