In the recent Qlik Sense 3.1 Service Release 2 I see that Qlik has introduced a new “Shared Persistence” storage model for multi-node installations. In this post, I will quickly explain what the “Share Persistence” model is and what that means for the installation process. I am not going to comment on Qlik’s reason for changing or the advantages/disadvantages of switching to the Shared model.
The old or “other” (I say other as it is still an option but I am not sure how long it will be available) architecture called the “synchronized persistence” storage model is where all Apps, Repository Content and shared files were physically replicated to each node. Whereas with the new Persistence storage model these Apps, Repository Content and Shared files are stored centrally and loaded into memory on the Rim nodes.
Setup and prerequisites
Apart from some initial set-up of the Folder share (where things are centrally stored), there isn’t too much that has changed for a basic installation.
Folder share Set-up (taken directly from Qlik Sense Install document)
The files that are persisted in a Qlik Sense deployment must be available to all nodes. They can be stored on any of the nodes in the cluster, or on a separate server.
- Create a folder for all file content to be shared, for example, c:QlikShare.
- Create a Windows share (like QlikShare) for this folder and set read/write access to this share for the service account that is used to run the Qlik Sense services.
This file share should now be accessible from all machines as \<servername>QlikShare
Qlik Sense Installation
I’ll take you through the screen shots (performed on version 3.1.4) you’ll notice not much has changed and the install process is as simple as ever. Note 3.2 has just been released and there may be some additional changes.
Choose storage model – Share persistence is the new way introduced by Qlik.
Standard License Jargon
New Cluster or Adding a Node to an existing Cluster
Shared persistence database details – this installation used the default local host however you are able to host the Postgres Database on another dedicated machine, this would require you to install and configure the DB separately prior to installing Qlik Sense.
Shared persistence storage details – the Root folder is the Share you created in the section above
Install Directory – I installed on D: although I found some logs etc are still found on the C: but I believe you can modify after installation.
Repository Database Superuser Password
Service Credentials – The credentials for running the Qlik Services, it should have Local Admin rights on the server.
Host Name (For self-signed certificates?)
There you have it Qlik Sense Installed with Shared Persistence, you’ll need to apply your license, user tokens, and potentially add to your server/proxy whitelist but that’s it and users should be able to use Qlik Sense without knowing that the underlying architecture has changed…
Helpful Qlik Sense 3.2 documentation (Note you may need a Qlik Community login).
You learn something every day.
Barry, Preventer of Chaos.
Barry blogs about how to stop chaos in your systems. Want to read more? Try SAS Viya: What is it? How can it help me? Or more from Barry.
We’ve got the best SAS Administrators in New Zealand ready to help you with your SAS environment.