Product Version: 6.8.x
Updated: Nov 28, 2018
For those familiar with enabling HA and setting up a ProfileUnity Console cluster, the process has not changed. We've only added a new reason why you may want to enable HA and clustering. The ProfileUnity Console needs to be online to fulfill license requests from the client tools. If the server is down, the users will get a message saying the license server is unavailable after they login or every hour if using AWS hourly-billing.
There is no concern of CPU usage or overloading a console with too many license requests, this is merely for the concern of users getting a "License Server Unavailable" message during a node-down situation.
- If you're running an older version of ProfileUnity that is NOT in a cluster configuration yet, then wait until after upgrading to 6.8.x before beginning this guide.
- N number of additional ProfileUnity Console servers, each installed fresh and NOT cloned from your original standalone instance. The total number of nodes in the cluster must be an ODD NUMBER. We recommend 3 nodes, but 5 or 7 are supported as well. Generally more than 3 isn't necessary.
- The number of nodes that can go down and have the cluster still operating normally depends on the total number of nodes in the cluster. Normal operation requires (at least) 2 nodes to be up at all times. So a 3 node cluster would allow for a single node failure, a 5 node cluster would allow 3 nodes to fail and so on. Two nodes is not a valid cluster configuration and will NOT result in the ability to lose a node and still function normally!
- Once you have installed the ProfileUnity Console software on each of the nodes, you only need to configure settings on your primary node. The others will inherit all settings/data once they are added to the cluster. You don't need to login to the secondary consoles at all.
- On your primary console node, navigate to the top-right menu with your login name, Administration screen, Clustering section toward the bottom.
- Put a check in the Enable High-Availability Mode box, and click Update at the top right of the screen. Wait for the system to switch into HA mode and for this screen to complete.
- Once that screen goes away, you will be redirect to the login page. Log back in and navigate back to the Administration screen, Clustering section again, you will click on the Add Node link in the top right of the Node Statuses box and enter the FQDN of a secondary node. Then repeat for the other(s) secondary node(s) as needed.
- Once complete, you should have all of your added nodes listed in the status with green/white check-marks.
- Next, we need to make sure that the clientsettings.xml file on your network path was updated by checking the date on the file. If not, we need to Download Client Settings or Deploy Client Settings to the correct location, either NETLOGON where your client tools/INI files are or another location previously specified in your ProfileUnity Computer GPO.
*** Make sure you're using the correct Communication Type needed for your environment, this example does not use VMDK FlexDisks/FlexApps or VHD ProfileDisks so the License Only option is selected.
- Last, we need to force the machines to pickup the new connection string from the new clientsettings.xml file. This only happens when the LwL.ProfileUnity.Startup.exe is (re)run. If you have this configured as a GPO startup script, then you should just need to refresh your pool(s) or reboot your physical machines. If this GPO is not in use, you will need to find another way to have this command re-run on your machines/pool(s) or master image template(s).
In order to verify that your machine client tools are now seeing/using all ProfileUnity Console nodes, on one of the machines you can navigate to C:\Windows\Temp\ProfileUnity and take a look in the most recent clientlicenseservice log file. You should see all the nodes listed there.