Procedure

  1. If you have multiple racks in your datacenter, stop SDDC Manager and LCM services on each additional rack. In a single rack scenario, the ISVM upgrade scripts stops the services so you can proceed to step 2.
    1. Using the root account, SSH to the rack's SDDC Manager VM.
    2. Type the following commands.

      service vrm-watchdogserver stop

      service vrm-tcserver stop

      service lcm-watchdogserver stop

      service lcm-init stop

      Leave this console window open.

  2. In a command line window, SSH to the 192.168.100.x IP address for the SDDC Manager VM on rack 1.
  3. Type the following.

    ls -lt /home/vrack/lcm/upgrade/vrm

    Note the upgrade ID displayed on the top row.

  4. Navigate to the following directory. upgrade_id is the upgrade ID you noted in step 3.

    cd /home/vrack/lcm/upgrade/vrm/upgrade_id/vrm-upgrade-vcf212-vcf213/isvm/scripts/

  5. Run the following command to update the isvm-upgrade.conf file.

    python isvm_upgrade_autoconf.py > isvm-upgrade.conf

  6. Run the following command to upgrade the ISVMs.

    ./isvm-upgrade.sh vm-patch isvm-upgrade.conf ../packages/isvm-vcf212-to-vcf213-upgrade.tar.gz | tee isvm-patch_upgrade.log

  7. After the ISVM upgrade is complete, press control-C to exit the console.
  8. Start services on each additional rack. SSH to each SDDC Manager VM and run the following commands.
    service vrm-watchdogserver start
    service lcm-watchdogserver start
  9. If the ISVM upgrade fails, follow the steps below.
    1. SSH to each ISVM and run the following commands.

      /etc/init.d/zkserver start
      /etc/init.d/ismd start
      /etc/init.d/ism-watchdog start
      /etc/init.d/cassandraserver start
    2. SSH to each SDDC Manager VM and run the following commands.

      service vrm-watchdogserver start
      service lcm-watchdogserver start
    3. Contact VMware Support and fix the error before proceeding with the upgrade. Your current environment will be functional even though the upgrade has not been completed.