7.3

Table Of Contents
Starting in vRealize Automation 7.3, you no longer need to manually start or stop the Manager Service on
each Windows server, to control which serves as primary or backup. Automatic Manager Service failover
is enabled by default in the following cases.
n
When you install vRealize Automation silently or with the Installation Wizard
n
When you upgrade IaaS through the administration interface or with the automatic upgrade script
When automatic failover is enabled, the Manager Service automatically starts on all Manager Service
hosts, including backups. The automatic failover feature allows the hosts to transparently monitor each
other and fail over when necessary, but the Windows service must be running on all hosts.
Note You are not required to use automatic failover. You may disable it and continue to manually start
and stop the Windows service to control which host serves as primary or backup. If you take the manual
failover approach, you must only start the service on one host at a time. With automatic failover disabled,
simultaneously running the service on multiple IaaS servers makes vRealize Automation unusable.
Do not attempt to selectively enable or disable automatic failover. Automatic failover must always be
synchronized as on or off, across every Manager Service host in an IaaS deployment.
If automatic failover does not appear to be working, see Upgrading from vRealize Automation 7.1 or 7.2
to 7.3 for troubleshooting tips.
Automatic vRealize Automation PostgreSQL Database
Failover
In a high availability vRealize Automation deployment, some configurations allow the embedded
vRealize Automation PostgreSQL database to fail over automatically.
Automatic failover is silently enabled under the following conditions.
n
The high availability deployment includes three vRealize Automation appliances.
Automatic failover is not supported with only two appliances.
n
Database replication is set to Synchronous Mode in vRA Settings > Database in the
vRealize Automation administration interface.
Usually, you should avoid performing a manual failover while automatic failover is enabled. However, for
some node problems, automatic failover might not occur even though it is enabled. When that happens,
check to see if you need to perform a manual failover.
1 After the primary PostgreSQL database node fails, wait up to 5 minutes for the rest of the cluster to
stabilize.
2 On a surviving vRealize Automation appliance node, open a browser to the following URL.
https://vrealize-automation-appliance-FQDN:5434/api/status
3 Search for manualFailoverNeeded.
Installing vRealize Automation
VMware, Inc. 133