Details

    • Bug
    • Resolution: Not a Bug
    • Minor
    • None
    • Lustre 2.8.0
    • None
    • CentOS 7.2

    Description

      When ever a reboot/shutdown command is issued from OS
      It's failing because LNET didn't shutdown properly

      Attachments

        Activity

          [LU-9665] LNET not shutting down properly

          Bhanu, Can you please share how you updated the script.

          ashehata Amir Shehata (Inactive) added a comment - Bhanu, Can you please share how you updated the script.
          pjones Peter Jones added a comment -

          Great - thanks

          pjones Peter Jones added a comment - Great - thanks

          Thanks,
          we updated the script and now it's shutting down properly
          You can close this ticket

          bhanuroot Bhanu (Inactive) added a comment - Thanks, we updated the script and now it's shutting down properly You can close this ticket

          I believe the order when shutting down should be to
          1. umount the FS
          2. lustre_rmmod
          3. reboot/shutdown

          There should be an lnet script

          /usr/lib/systemd/system/lnet.service 
          

          That script should do the above steps.

          It'll be a matter of making sure that this script is executed (IE lnet service stopped) on shutdown before the OFED modules are unloaded.

          ashehata Amir Shehata (Inactive) added a comment - I believe the order when shutting down should be to 1. umount the FS 2. lustre_rmmod 3. reboot/shutdown There should be an lnet script /usr/lib/systemd/system/lnet.service That script should do the above steps. It'll be a matter of making sure that this script is executed (IE lnet service stopped) on shutdown before the OFED modules are unloaded.

          Is there a recommended init script to shutdown lnet ??

          bhanuroot Bhanu (Inactive) added a comment - Is there a recommended init script to shutdown lnet ??

          It appears that OFED is being shutdown before LNet. This would cause LNet to get stuck. So in your shutdown procedure LNet should be shutdown before OFED.

          ashehata Amir Shehata (Inactive) added a comment - It appears that OFED is being shutdown before LNet. This would cause LNet to get stuck. So in your shutdown procedure LNet should be shutdown before OFED.

          People

            wc-triage WC Triage
            bhanuroot Bhanu (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: