Upgrade your vCenter 6.5 HA environment

As discussed in my previous post here, you can easily setup vCenter HA to provide a decent (not the best and hopefully this will improve) RTO of around 4 minutes for a fail over of your vCenter server.

So now that you have vCenter HA configure, how do you patch or upgrade this environment.  In a single vCenter Server instance the upgrade is really straight forward.

  • Login to the VAMI
  • Before starting the upgrade, take a File based Backup of the vCSA, using the backup utility in the VAMI.
  • Select Update
  • Select Check Updates -> Check Repository (if you have internet access)
    • Otherwise download the software and mount the ISO to the CD/DVD drive.
  • View Available Updates
  • Screen Shot 2017-10-10 at 10.30.25 AM.png
  • Select Install Updates -> Install All Updates

 

For a vCenter HA the steps are bit more complicate since we will use the software-packages utility from the appliance shell which requires us to SSH into the three nodes in a sequence and use manual failover so that we always patch the non-active node.  Below are my quick step by step notes for the upgrade process:

  • There are multiple ways to use the software-package utility:
    • Use the default repository
    • Use a local repository by attaching the ISO to the vCenter Server appliance.
    • Use a remote repository by using a custom repository URL that points to a local webserver in your environment to retrieve the file.
  • In my case I downloaded the vCenter Server Appliance patch ISO from “https://my.vmware.com/group/vmware/patch” and attached the file to the CD/DVD drive of the vCSA.
  • Before I start the upgrade I perform the following tasks:
    • Put the vCenter HA cluster in maintenance mode
    • Make sure SSH is enabled in the vCSA VAMI
    • For each node, I open the console and mount the patch ISO to the CD/DVD drive.
    • Take a File based Backup of the vCSA, using the backup utility in the VAMI.
  • Run the upgrade first on the Witness Node
    • First off SSH into the active vCSA node
      • From the active vCSA node, SSH into the witness node and make sure you are in the appliance shell by running:
        • “appliancesh”
        • Run: “software-packages install –iso”
        • Press Enter way to many times
          • Type yes and press Enter
        • When upgrade is completed, reboot the server
          • “shutdown reboot -r patching”
        • Exit the SSH session
  • Now run the upgrade on the Passive Node
    • First off SSH into the active vCSA node
      • From the active vCSA node, SSH into the passive node and make sure you are in the appliance shell by running:
        • “appliancesh”
        • Run: “software-packages install –iso”
        • Press Enter way to many times
          • Type yes and press Enter
        • When upgrade is completed, reboot the server
          • “shutdown reboot -r patching”
        • Exit the SSH session
  • Log out of the active vCSA node
  • Wait for the nodes to shows status up after reboot.
  • Initiate a vCenter HA failover manually
    • Login to Web client
    • Select the vCenter server -> Configure -> Settings -> vCenter HA
    • Click Initiate failover
    • Click Yes to start the failover
      • Make sure to select performing synchronization first
  • Now lastly run the upgrade on the new Passive Node
    • First off SSH into the new active vCSA node
      • From the active vCSA node, SSH into the passive node and make sure you are in the appliance shell by running:
        • “appliancesh”
        • Run: “software-packages install –iso”
        • Press Enter way to many times
          • Type yes and press Enter
        • When upgrade is completed, reboot the server
          • “shutdown reboot -r patching”
        • Exit the SSH session
  • Optional: Perform another vCenter HA failover manually back to the original vCSA node.
  • Exit vCenter HA maintenance mode
    • Login to Web client
    • Select the vCenter server -> Configure -> Settings -> vCenter HA
    • Click Edit
    • Select “Enable vCenter HA”
    • click OK

Patching of all the vCenter HA nodes should now be completed.

Food for though: This process is quite involved and I wonder, depending on company policy, would it not be easier to just remove vCenter HA, upgrade the single vCSA node through the VAMI and then configure vCenter HA again? It takes way less time and much simpler process. Let me know what you think.

 

 

 

Installing your vCenter Server 6.5 HA environment

vSphere 6.5 finally brought us vCenter High Availability which provides failover/RTO such that users can continue with their work in around 2 minutes through
API clients and in my observations around 5ish minutes through UI clients, awesome right!

vCenter HA is really simple to configure through the Wizard with a basic and advanced option, with the following requirements:

  • vCenter Server 6.5 Appliance with internal or external PSC
  • vCenter Server 6.5 Appliance with no snapshot and CD/DVD drive set to Client Device.
  • Cluster with at least 3 ESXi hosts that this vCSA is both managing and running on will provide Basic installation option
  • If your vCSA is managed by another vCenter Server that is not part of the same SSO domain, or running on a host that is does not manage, you will have to use the Advanced installation option.
  • A port group on ESXi for the private HA network. Optionally, you can have a dedicated vSwitch if network isolation is a requirement.
    • The HA private network must reside on a different subnet than what is used for management.
  • Network latency on the HA network must be less than 10ms.
  • No gateway for the HA network must be specified when configuring the nodes manually during advanced installations.
  • Storage:
    • Datastores for peer and witness cannot resides in a storage cluster.
    • Will get warning if you put nodes in same datastore

When vCenter HA is enabled, a three-node vCenter Server cluster (Active, Passive, and Witness nodes) is deployed, which consists of one IP address for management and three private IP addresses for the HA nodes.

  1. Open vCenter Server 6.5 web client (HTML5 client does not yet support vCenter HA).
  2. Select the vCenter server in Navigator
  3. Select Configuration -> Settings -> vCenter HA
  4. Screen Shot 2017-10-26 at 9.30.46 AM.png
  5. Click Configure

Continue reading

vCenter Server Web Client crash with latest Flash 27.0.0.170

Within the last couple of days Google ran an automatic update of Flash, which is causing havoc in the all-things-flash world.

Since Google manages the upgrade of the flash you are kind of at their mercy on this one, until all applications finally learn to stay as far away from flash as possible! Can only hope!

You can however disable the automatic update feature for flash by setting the registry HKLM\SOFTWARE\Policies\Google\Update\AutoUpdateCheckPeriodMinutes to 0.

My fix was pretty straight forward and I only had to delete the latest release folder from the path “%LocalAppData%\Google\Chrome\User Data\PepperFlash” (windows) or “~/Library/Application Support/Google/Chrome/PepperFlash” (MAC).

In my case the older version was still available in the same folder so I only had to delete the latest and restart google but if do not have the older version you can download it here.

From other users responses it seems that the version 27.0.0.130 and older still works but newer version will crash on both your Google Chrome and Firefox browsers.

Update:  VMware has release a KB here

 

 

Migration of vCenter Server using VMware Migration Assistant

The migration assistant can tackle a couple of different vCenter Server configurations:

  • vCenter Server 5.5 or 6.0 with an embedded vCenter Single Sign-On instance on Windows
  • vCenter Server 5.5 instance on Windows with external SSO
  • vCenter server 6.0 instance on Windows with external PSC

Depending on your current configuration, the migration process will change based using an internal or external SSO/PSC and where VUM is installed.

  • If VUM is installed on a separate Windows server, other that your Windows vCenter Server instance then you MUST to run the migration assistant on this server FIRST!  The VMware Migration Assistant facilitates the migration of the Update Manager server and database to the vCenter Server Appliance 6.5.
  • If embedded SSO/PSC, then you run the Migration Assistant on the source vCenter Server
  • If external SSO/PSC, then you run the Migration Assistant on the source SSO/PSC first and then run the Migration Assistant on the source vCenter Server
    • During the migration process, make sure to leave the migration windows open

You run the VMware Migration Assistant on your source vCenter Server, SSO/PSC or VUM it will perform the following tasks:

  • Discover the source deployment type.
  • Run pre-checks on the source.
  • Report errors that must be addressed before starting the migration.
  • Provide information for the next steps in the migration process.

Here are my step by step instructions for vCenter Server 5.5 with embedded SSO, no VUM:

  1. Download and Mount the vCenter Server Appliance Installer ISO file to the Windows vCenter Server.
  2. Start the migration assistant on the source, depending on your environment configuration
    1. Open the installer folder on CD/DVD drive and copy the  “migration-assistant” folder to your source windows server running the vCenter Server
    2. Open the copied folder on local drive and run “VMware-migration-assistant.exe”
      1. This will run the pre-update checklist and prompt if anything needs to be fixed.
      2. Make sure this window stays open during the whole migration process until completion.
  3. Launch vCenter server appliance UI installer
    1. Click Migrate
    2. Click Next on Introduction
    3. Accept EULA and click Next
    4. Connect to the source server
      1. Enter Windows server FQDN or IP address
      2. Enter migration assistant port number
        • If you are changing networks and have a firewall in place then pick a port that is open and available.
      3. SSO username
        • administrator@vsphere.local
      4. SSO Password
      5. Click Next
    5. Accept Thumbprint
    6. Appliance deployment target
      1. Enter ESXi or vCenter Server name for targer
      2. HTTPS port
      3. Username
      4. Password
      5. Click Next
    7. Accept Thumbprint
    8. Select Folder to place the new vCSA
    9. Select the compute resource
    10. Target appliance VM
      1. Enter VM name
      2. Enter new root password
    11. Select deployment size
    12. Select datastore
    13. Configure the temporary network identity which is used to copy the data, this will be removed after migration is completed since the target vCSA server will get the network identity of the source vCenter server.
      1. Pick network port group
      2. Enter temp IP address
      3. Enter subnet
      4. Enter GW
      5. Enter DNS servers
      6. Click Next
    14. Verify detailed entered are correct!
    15. Click Finish
    16. Wait for the migration to complete.
      • If the migration fails for any reason your can easily roll back by powering off the target vCSA and then just power on the source vCenter Server and its components.
  4. Now for stage two, which copies the data from the source vCenter to new vCSCA.
    1. On Introduction click next
    2. Connect to Source vCenter should complete automatically
    3. Join AD
      • Add username and password for user that can add computer to domain
    4. Select the best migration data options with the following 3 being available. (pretty cool that it provides the size of the migration for each option as well)
      • Just the configuration
      • Configuration, Events and Tasks
      • Configuration, Events, Tasks and Performance metrics
    5. Select if you want to join CEIP and click Next
    6. Review your settings, check the box to verify you have a backup of the vCenter server as well as the database.  Click Next
    7. Click OK on prompt that warns you that the source vCenter Server will be shut down once the network configuration is enabled.
    8. Wait for the data migration to complete.

When successful, your source vCenter Server should now be shut down and brand new shiny vCenter Server appliance should be started in its place. The best part of all you are also upgraded to vCenter Server 6.5! awesome.

 

 

vCenter Server 6.5 error: Cannot start content-library service

I recently discovered a problem with my vCenter Server after an upgrade from 6.0 to 6.5 of the vCenter Server Appliance, which runs in linked mode.

When I tried to take a backup of the vCenter Server through the VAMI I got an error “Invalid vCenter Server Status: All required services are not up! Stopped services: ‘content-library’.

Troubleshooting:

My first attempt was trying to start the service

  • I logged into the vCenter Server web client and went to administration -> Deployment -> System Configuration.
  • Double click on the the vCSA node
  • Select services
  • Select Content Library service, which under summary showed stopped.
  • Click in Actions -> Starts
  • Service failed to start with very cryptic error “An error occurred while starting service ‘%(0)s”

I also ssh into the vCSA and tried to start the service, and see the following error when trying to start the service:

Error executing start on service content-library. Details {
“resolution”: null,
“detail”: [
{
“args”: [
“content-library”
],
“id”: “install.ciscommon.service.failstart”,
“localized”: “An error occurred while starting service ‘content-library'”,
“translatable”: “An error occurred while starting service ‘%(0)s'”
}
],
“componentKey”: null,
“problemId”: null
}
Service-control failed. Error {
“resolution”: null,
“detail”: [
{
“args”: [
“content-library”
],
“id”: “install.ciscommon.service.failstart”,
“localized”: “An error occurred while starting service ‘content-library'”,
“translatable”: “An error occurred while starting service ‘%(0)s'”
}
],
“componentKey”: null,
“problemId”: null
}

Since the error message did not help much I went to search on VMware community and found other users having issues with creating content libraries after upgrade and VMware support pointed them to a KB article.

The fix:

The solution comes from KB 2151085 , however I am on version 6.5 not 6.5U1, as well as the OVF stuff was not applicable to my particular situation but the steps provided did resolve my problem and I was able to start the content-libary services.

I compared the files with a vCenter 6.5 environment that is working and found that the ts-config.properties file had the owner and group set to root:root and should be set to content-library:cis.  This is an easy fix and also solved the problem for me.   “chown content-library:cis ts-config.properties”

 

 

 

Upgrading vROPS 6.x to 6.6.1

With all the new goodies in 6.6, especially the new HTML5 UI based on the Clarity design System, who can resist the upgrade to vROPS 6.6. Release notes for everything that is new can found here.

From an upgrade standpoint, vROPS has always been an interesting, but simple process with both the OS and application that requires separate updates.  The OS update is required for update RPMs for things like database and gemfire updates that the new vROPS application relies on.  My step by step upgrade guide below:

  1. Download the OS update and Product update files from my.vmware.com
    • OS PAK file:  vRealize_Operations_Manager-VA-OS-xxx.pak
    • Application PAK file:  vRealize_Operations_Manager-VA-xxx.pak
  2. Make sure that all the solutions you have installed has a version available that is compatible with the new vROPS release.
  3. If you customized any default alert definitions, symptom, recommendations, Policy Definitions, Views, Dashboards, Widgets and Reports in the previous version, make sure you clone it first.  When you upgrade vROPS, it is important that you upgrade the current versions of content types that allow you to alert on and monitor the objects in your environment.  It is a good practice to always clone first before customizing content.
  4. Before starting the upgrade, create a snapshot of the each of the nodes in the cluster.
    1. Login to vROPS admin
    2. Under system status click Take Offline
    3. Enter reason and click OK
    4. When Cluster status shows offline for all nodes, go ahead and take a snapshot of each
  5.  Before starting the upgrade, I also recommend taking a backup of all the nodes simultaneously by using your existing backup solution.
  6. First off we will update the Virtual Appliance OS:
    1. Login to the master vROPS node administrator interface
    2. Select Software Update
    3. Click Install a Software Update
    4. Browse the OS update PAK file
      • vRealize_Operations_Manager-VA-OS-xxx.pak
    5. Check the box “Reset Default Content”
      • As mentioned above make sure you have cloned all your customized content!
    6. Click Upload
    7. When completed click Next
    8. Accept EULA click Next
    9. Click Next
    10. Click Install
    11. This will update the OS on the Virtual Appliances and restart them.
  7. Secondly we will perform the vROPS product update:
    1. Login to the master vROPS node administrator interface
    2. Select Software Update
    3. Click Install a Software Update
    4. Browse the application update PAK file
      • vRealize_Operations_Manager-VA-xxx.pak
    5. Check the box “Reset Default Content”
      • As mentioned above make sure you have cloned all your customized content!
    6. Click Upload
    7. This will update the vROPS application on the Virtual Appliances
  8. Lastly, if you have any additional content packs installed, go ahead and upgrade them.

VMware is definitely making awesome improvements in all their products and has come a long way in helping out VMware admins with their daily management tasks.

SovLabs: Upgrading your software

SovLabs isn’t just a vRA plugin, it’s enterprise software that extends the capabilities of your vRealize Automation environment providing you with that end-to-end solution you have been craving for.  As with any other enterprise software they periodically provide new patches and releases and with SoLabs that is no different.

The new 2017.3.x was released in August and provides some awesome new modules:

  • Men & Mice DNS and IPAM
  • SolarWinds DNS
  • Backup as a Service
    • Automate policy-driven backups and provide self-service VM and file-level recovery for –
    • Cohesity
    • Rubrik
    • Veeam
  • SovLabs VM tagging
    • Drive rich metadata using VM tags and categories
  • SovLabs Property Toolkit
    • Manage your existing custom properties on VMs with the SovLabs Template Engine
  • ServiceNow Support for Jakarta
  • Puppet support for 2017.1
  • VMware Tools connection
    • Connect to Windows/Linux servers can now be done through VMware Tools which removes the requirement for WinRM, CygwinSSH or WinSSHD to be installed.  This is huge!
  • As a customer you can sign up under the self-service portal and view the detailed release notes here:

So how do we go about upgrading SovLabs to the latest version?

Step by step guide to upgrading from 2017.2.x to 2017.3.x.  (there are some additional steps if you are upgrading from <= 2017.1.x so please contact SovLabs support) 

  1. First off we want to create a backup of the vRO package
    1. Login to vRO Client
    2. Click Design
    3. Click on the package tab
    4. Click on the package icon on right hand side menu bar
    5. Enter name “com.sovlabs.backup.resources”
    6. Edit the newly create package, click on the pencil icon on the right hand side menu bar
    7. Click the Resources tab
    8. Click the Folder + icon
    9. Expand the Library folder,  select the SovLabs folder
    10. Click on the Select button
    11. Once loaded, click save and close
    12. Right click the saved package and click export package
      1. Create a folder called sovlabs under downloads
      2. leave the rest of settings as default
    13. Save to your local system
    14. Now, lets save the old SovLabs Plugin:
      1. Use WinSCP and login as root to vRO appliance
      2. Go to directory /var/lib/vco/app-server/plugins
      3. Save the o11nplugin-sovlabs.dar to your local file system in same sovlabs folder created earlier under download.s
    15. Done!
  2. We need to update the vRO Heap size
    1. If you have done this before then you can skip this step but this is needed to install the larger sized SovLabs module file into vRO otherwise the appliance might run out of memory during install/upgrade.
    2. Remember if you a vRO cluster, then you have to perform the steps on both server
    3. SSH into vRO appliance with user root
    4. Run # vi /var/lib/vco/configuration/bin/setenv.sh
    5. Find the #MEM_OPTS section
    6. Replace the -Xmx512m \ with -Xmx768m \
    7. Save the file
  3. Delete all SovLabs license keys
    1. Login to vRA tenant
    2. Click on Items tab -> SovLabs vRA Extensibility modules -> SovLabs License
    3. For each SovLabs License item listed
      1. Select Actions -> Delete License
  4. Download the SovLabs plugin
    1. Talk to SovLabs support about getting the software downloaded.
  5. Install the plugin into vRO appliance
    1. Login to controlcenter
      1. https://<vroserver&gt;:8283/vco-controlcenter
    2. Select Plug-Ins -> Manage Plug-ins
    3. Click Browse
    4. Select the plugin
    5. Accept EULA
    6. Click on Install
    7. Accept the EULA
    8. Restart the vRO server
      1. On the Home page, click on the Startup Options icon
      2. Click on Restart
      3. Wait for vRO to restart successfully
    9. Log back in to the vRO configuration page
    10. Click on the Manage Plug-Ins icon
    11. Verify that the installed plugin is listed among the vRO plugins
    12. Now if you have a clustered vRO 7.2 and above, then the plugin should sync but I have seen some problems with 7.2 so follow these steps
      1. Perform a full reboot on primary so that the pending and active config fingerprint ID match.
      2. Then push the config to the other standby node
      3. It will need to rebooted which it often will not do so make sure you perform this step yourself.
      4. Verify that Synchronization state shows synchronized and verify the version of the plugin on both active and standby nodes.
  6. Login to the vRO Client and run the configuration
    1. Click on Design mode
    2. Click on WorkFlow tab
    3. Right click vRO workflow, “SovLabs/Configuration/SovLabs Configuration”
    4. Select Start Workflow
    5. The SovLabs Configuration workflow only needs to be run on one vRO in a clustered environment
      1. Select yes to accept the EULA
      2. Click Next
      3. Select the appropriate tenant and business group
      4. Create SovLabs vRA Catalog Service? = No
      5. Publish License Content? = No
      6. Click Next
      7. Upgrade existing SovLabs vRA content? = Yes
      8. Click Next
      9. Install or Update SovLabs workflow subscriptions (vRA7.x)? = Yes
        1. *Enables vRA to call vRO during machine lifecycles
      10. Click Submit
      11. Verify that the SovLabs Configuration workflow completed successfully
  7. Lastly, let’s verify the SovLabs Plugin in vRA
    1. Select Catalog tab
    2. Verify that Add license -> SovLabs Modules catalog exists
  8. Now lets install the new license key for 2017.3.x
    1. This process has also been drastically simplified with a single license key which will license all modules, where previously this was done one at a time.
    2. Select Catalog tab -> SovLabs vRA Extensibility Modules -> Add license – SovLabs Modules
    3. Copy the text from license file and paste into field
    4. Click Submit
    5. Verify all catalog tab -> SovLabs vRA Extensibility Modules that all catalogs are available.
  9. If you ever need to roll back then follow the steps in the document provided by SovLabs:
    1. https://s3.amazonaws.com/docs.sovlabs.com/vRA7x/guides/SovLabs_BackupRestore-vROPackage.pdf