I haven't tested if there is a feature in the new release which allows me to define a node as secondary in QMC but the STT video was leading me in that direction. The other option is to have a Windows failover cluster for the central node but this means we will waste an entire server.
(2) Breakdown of external server for shared folder and repository
Windows failover clustering for the file server serving the shared folder is recommended. The disk space for the shared folder itself should be on SAN controller ideally.
And for the repository server which is a Postgresql implementation, you can refer to the HA and replication options available for Postgresql. The disk space for the Postgresql db should be on a SAN controller ideally for HA.
The Qlik Sense environment can also be hosted in a virtualized environment and in such cases the HA and DR options offered by the hypervisor environment can also be utilized
Re: How to achieve High Availability in Qlik Sense?
Secondary paragraph: "Qlik Sense June 2017 and later versions do not support the synchronized persistence model. To upgrade from Qlik Sense 3.1 SR2 or later to Qlik Sense June 2017 or later and migrate from a synchronized persistence model to a shared persistence "