Note: An upgrade of a Blueprint server (Collector/Analyst) on a cluster cannot be rolled back. As a precaution, certain resources should be backed up before upgrading. This enables a manual upgrade if the upgrade fails for any reason.

The following steps should be carried out when upgrading a clustered 3.5 R2 or 4.1 Analyst/Collector to 4.2.

  1. Using the Cluster Administrator, take the “Blueprint TaskMaster” and “Blueprint ComTaskMaster” resources offline. Temporarily change the properties of the “Print Spooler” and “TCP/IP Print Server” resources – on the Policies tab, uncheck the option to fail over if restart is unsuccessful. Ensure all nodes are showing “Up” status before proceeding.
  2. Run the upgrade on the first, active node. When the Server Configuration Tool (SCT) appears, click the Apply button, and then click the Close button.
  3. Open the Services Manager on the active node. For both the “Pharos Systems TaskMaster” and “Pharos Systems ComTaskMaster”, set the startup type to “Manual” and configure the services to use the correct domain user account.
  4. Fail over the node. Keep the “Blueprint TaskMaster” and “Blueprint ComTaskMaster” resources offline.
  5. Run the upgrade on the second, now active, node. When the SCT appears, click the Apply button, and then click the Close button.
  6. Using the Cluster Administrator, restore the properties of the “Print Spooler” and “TCP/IP Print Server” resources – on the Policies tab, check the option to fail over if restart is unsuccessful.
  7. Open the Services Manager on the active node. For both the “Pharos Systems TaskMaster” and “Pharos Systems ComTaskMaster”, set the startup type to “Manual” and configure the services to use the correct domain user account.
  8. Add the two new services - Pharos Systems Job Service and Pharos Systems Event Log Service to the resource group as Generic Service resources.
    1. Blueprint Job Service
      1. Dependencies: Blueprint Event Log Service, Physical Disk, Network Name, Spooler
      2. Service Name: Pharos Systems Job Service
      3. Use Network Name: True
      4. Affect the Group: False
    2. Blueprint Event Log Service
      1. Dependencies: Physical Disk, Network Name
      2. Service Name: Pharos Systems Event Log Service
      3. Use Network Name: True
      4. Affect the Group: False
  9. Set the startup type to “Manual” for the two new services, and configure the Pharos Systems Job Service to use the correct domain user account.
  10. Using the Cluster Administrator, bring both of the Blueprint resources online.
  11. Click the Test button on the Server Configuration Tool.