A Provider VM outage will block all Data Management for VMware Tanzu operations for users of any role, except:
- Scheduled database backups.
- Replication between Primary and Read Replica databases.
It is your responsibility to choose a tool and configure it to monitor the Data Management for VMware Tanzu web endpoint.
Detecting a Failure
The important components in a Provider VM include:
- Monitoring Service
- Provider API
- Internal vPostgres database
- RabbitMQ
Extended downtime or degradation of any of these components will affect the reliability of your Data Management for VMware Tanzu installation.
Evaluating the Failure
How do you determine when an issue is severe enough to manually invoke failover and promote a Standby Provider?
- Evaluate the health of the Provider VM.
- Evaluate the connectivity of the network interfaces located on the Provider VM.
- Evaluate the health of the Data Management for VMware Tanzu services running on the Provider VM.
Based on the information that you gather, you may choose to failover and promote a Standby Provider, or to repair the existing Primary Provider.
Choosing Failover
If you choose failover, you must identify a Standby Provider to promote.
After you ensure that the Standby Provider that you choose for failover is healthy and is synchronized with the Primary, promote this Standby Provider.
Choosing Repair
If you choose to repair the existing Provider VM, perform the required tasks, and then verify that the Provider and the HA cluster are functioning correctly.
