Update 1607 for Cloud Platform System (CPS) Premium

To install update 1607 for CPS Premium, follow these steps. Any references to the CPS Premium Administrators Guide refer to the Update 1607 version that's provided by your account team.


Note This update procedure assumes that you have already installed Update 1603.

Step 1: Update network switch firmware

Follow the steps in the "Update network switch firmware" section of the 1607 version of the CPS Premium Administrators Guide that was provided by your account team. Don't use instructions from a previous version of the Administrators Guide.


IMPORTANT: For Update 1607, the AFM update must be done through the console. Updating AFM through the GUI will corrupt the database and AFM will lose connectivity to the switches.


The download location of the AFM files for Update 1607 is:

http://poweredgec.dell.com/cps/CPS_P_1607/Tools/AFM

These are the recommended firmware versions that switches should run after the update:

Step 2: Update serial port concentrator firmware

Follow the steps in the "Update serial port concentrator firmware" section of the CPS Premium Administrators Guide that was provided by your account team.


Download locations are:

• Avocent ACS 6000 Series:
  http://download2.emersonnetworkpower.com/SerialACS/ACS6000/ACS6000_v3.3.0.10/avoImage_avctacs-3.3.0-10.zip
• Digi CM 48:
  http://ftp1.digi.com/support/firmware/80007070_U.bin

These are the recommended firmware versions that the serial port concentrator should run after the update.

Step 3: Install the VMM hotfixes for VM placement

Notes

• If you have already installed the VMM hotfix for virtual machine placement that was included in 1606, you don't have to install it again, and you can skip this step.
• The highly available VMM clustered role (
  Prefix-HA-VMM) has two nodes,
  Prefix-VMM-01 and
  Prefix-VMM-02. In the instructions, we refer to these as Node1 and Node2.
  

To install the VMM hotfixes, follow these steps:

1. For the new VMM placement hotfix, follow these steps:
   
   1. From the update file drop location, copy the Placement_HF_UR9.exe file to a folder on a Console VM, such as C:\VMMHotfix. (If you want, you can use the same folder as for the previous hotfix.)
      
   2. Double-click the Placement_HF_UR9.EXE file, review the EULA, and then select Yes to accept the terms.
      
   3. Select a folder to store the extracted files. For example, select C:\VMMHotfix. Then, select OK.
      
2. Determine the passive VMM node. To do this, open a Windows PowerShell ISE session, and then run the following script:
   

   $prefix = $ENV:COMPUTERNAME.split('-')[0]
   $VmmServerName = "$prefix-HA-VMM" 
   $vmmServer = Get-SCVMMServer -ComputerName $VmmServerName 
   $activeNode = $vmmServer.ActiveVMMNode 
   $passiveNodes = @() 
   $vmmServer.FailoverVMMNodes | ForEach-Object { if($_.ToLower() -ne $activeNode.ToLower()){ $passiveNodes += $_ } } 
   Write-Host "Active Node: $activeNode" -ForegroundColor GreenWrite-Host "Passive Node: $passiveNodes" -ForegroundColor Green 

   This script returns the server name of the passive VMM node. (In our example, we assume that Node2 is the default passive node.)
   
3. In File Explorer, browse to the following folder on the passive node:
   

   \\<Prefix>-VMM-0#>\c$\Program Files\Microsoft System Center 2012 R2\Virtual Machine Manager\bin 

4. Make a backup copy of the following files in the \bin folder:
   
   Engine.Placement.ResourceModel.dll
5. In the VMM console, determine which host (in the management cluster) the passive VMM node is running on.
6. Open Hyper-V Manager, connect to the management cluster host that you identified in the previous step, and then connect to the passive VMM node.
7. On the VMM node, type powershell to open an elevated Windows PowerShell session, and then run the following command:
   

   Stop-Service SCVMMService
   Stop-Service SCVMMAgent 

8. Verify that the services have stopped. To do this, run the following command:
   

   Get-Service SCVMMService
   Get-Service SCVMMAgent 

   Note If you're prompted to close the System Center Management Service Host process, select Ignore.
   
9. On the Console VM, browse to the following folder on the passive node:
   

   \\<Prefix>-VMM-0#\c$\Program Files\Microsoft System Center 2012 R2\Virtual Machine Manager\bin 

10. In the \bin folder, replace the following file by using the new version of the files that you extracted from the hotfix package:
    
    Engine.Placement.ResourceModel.dll
11. On the passive VMM node, run the following command to start the VMM Agent service:
    

    Start-Service SCVMMAgent 

    Note The SCVMMService service doesn't start if the passive VMM server node isn't active. The SCVMMService service starts only when the node becomes the Active node. This behavior is by design.
    
12. In Failover Cluster Manager, start a failover. This will set Node1 as the new passive node and set Node2 (which has already been updated) as the active node.
13. Open Failover Cluster Manager, and then connect to the
    Prefix-CL-VMM cluster.
    
14. Select Roles.
    
    Note The Roles pane displays the active node in the Owner Node column.
    
15. Press and hold (or right-click) the active node, point to Move, and then select Select Node.
    
16. Select the other node, and make sure that the status changes to Running for the new active node.
    
    Note The change may take several seconds to occur.
    
17. Follow steps 512 to update the VMM file on the new passive node (in this example, Node1).

If you must revert the update, follow these steps:

1. On the passive VMM node, stop the SCVMMAgent service.
   
2. Replace the files under your Virtual Machine Manager Install directory with your backup files.
3. Start the SCVMMAgent service.
   
4. In Failover Cluster Manager, start a failover of the Prefix-HA-VMM clustered role.
   
5. Repeat steps 1-4 on the new passive node.

Step 4: Prepare the package

Follow the steps in the "Prepare the patching environment" section of the CPS Premium Administrators Guide that was provided by your account team.


Extract both the 1607-0 and 1607-1 updates now, and then put them into different folders on the Infrastructure share.


Important Don't start the update process. Instead, run the Health Check that's described in "Step 5: Run the Patch & Update Health Check".

Step 5: Clean up the WSUS server

To clean up the WSUS server, follow these steps:

1. On the Console VM, open the Windows Server Update Services console.
   
2. Press and hold (or right-click) Update Services, select
   Connect to Server, and then connect to the WSUS VM (
   Prefix-SUS-01).
   
3. In the navigation pane, expand
   Update Services > [WSUS Server] > Updates, and then select All Updates.
   
4. In the All Updates pane, select Any except declined in the Approval list.
   
5. In the Status list, select Any, and then select Refresh.
   
6. Select all updates.
7. Press and hold (or right-click) the selection, and then select Decline.
   
8. In the navigation pane, expand the server name, and then select Options.
   
9. In the Options pane, select Server Cleanup Wizard.
   
10. Select all check boxes except the Computers not contacting the server check box.  
11. SelectNext.
    

Step 6: Run the Patch & Update Health Check and fix discovered issues

In the "Update the computers" section of the CPS Premium Administrators Guide, complete "Step 1: Run a health check and fix any discovered issues." This includes functionality to check for and disable any running backup jobs.


Important Don't start the update process. Instead, run the Health Check, fix any discovered issues, and stop any running backup jobs.

Step 7: Run the P&U update package 1607-0

In the "Update the computers" section of the CPS Administrators Guide, follow "Step 2: Apply the P&U update package" to apply Update 1607-0.


Note Starting with Update 1603, the P&U (Patch and Update) engine automatically runs a health check as part of the update process. You can control what happens if critical Operations Manager alerts are discovered. To do this, change the value of the ScomAlertAction parameter. For more information, see the Administrators Guide.


Because of the size of this package, estimates for deployment are 1218 hours.


If you don't intend to apply 1607-1 immediately, remember to enable DPM agents if you disabled them earlier (as described in the Administrators Guide).


Also, if you don't intend to apply 1607-1 immediately, follow the steps in "Step 3: Post-update clean up" of the Administrators Guide after the updating is completed.

Step 8: Run the P&U package 1607-1

Follow the same steps for installing the 1607-1 package. This can be scheduled right after the 1607-0 package, or at a later time depending on your maintenance timeframe requirements.


Because of the size of this package, estimates for deployment are 36-48 hours (per rack).


After the patching process is complete, remember to enable DPM agents if you disabled them earlier (as described in the Administrators Guide). Also, follow the steps in "Step 3: Post-update clean up" of the Administrators Guide after the updating is complete.

Run an optional compliance scan

To run a compliance scan, pass the following flag:

\\SU1_InfrastructureShare1<CPSPU FolderName>\Framework\PatchingUpgrade\Invoke-PURun.ps1 -PUCredential $cred -ComplianceScanOnly 

The compliance scan output is written to the following location to which the update package was extracted. For example, the following output is written:

"PURoot"\MissingUpdates.json