Three Workspace one nodes per site and have AVI load balancer with external Database.
Here are the steps for testing that failover:
- Change the LB to point to the DR site.
- You will need to clear caches in the primary datacenter by following the steps listed here: https://docs.vmware.com/en/VMware-Workspace-ONE-Access/21.08/workspace_one_access_install/GUID-A794E38F-974E-48B9-8BFF-3737472BC1BB.html
- Restart all connector services from both the primary and DR site
- Stop the horizon-workspace service on the primary nodes. As we can only have 1 read/write cluster on at a time.
- Promote the DR site to the primary by editing: /usr/local/horizon/conf/runtime-config.properties in the following ways on each of the nodes:
- read.only.service=true to read.only.service=false
- cache.service.type=ehcache to cache.service.type=rds
- Restart the horizon-workspace server on each node in DR.
- restart the connectors again (to establish a new s-channel
- We should no longer see that banner in the console.
- Start the horizon-workspace service on each of the primary datacenter’s nodes.
- Change LB back to primary datacenter.
- Change back the DR site to read-only= true
- Change back the DR site to cache.srevicetype=ehcache
- Restart the horizon-workspace service in DR.
- Restart connectors in DR & primary.