VCP7-CMA – Objective 3.2

Disclaimer: These are my notes from taking the 2V0-731 exam. If something doesn’t make sense, please feel free to reach out.

The goal of this objective is to understand how to deploy vRA.

Objective 3.2 – Install an Enterprise Deployment

References

An Enterprise deployment involves distributed components and load balancing to give both scalability and availability.

With a load balanced deployment, having multiple web server instances & vRA appliances active during installation can cause the deployment to fail.

For distributed site clusters latency of less than 5ms is required, with a minimum bandwidth of 1Gb/s.

The appliance database is automatically clustered. For the IaaS database, a dedicated SQL server is recommended. To configure for HA, AlwaysOn is not supported due to it’s dependancy on MSDTC, Windows failover clustering is supported. AlwaysOn with SQL Server 2016 is supported in vRA 7.2.x but the exam is based upon 7.0.x.

For maximum performance, configure the proxy agents in the same data centre as the endpoint. Locate the DEMs close to the Model Manager.

External vRO is recommended for multiple tenants if you need to ensure total isolation.

Each IaaS server requires a management agent to be installed. If the primary vRA appliance fails, these will all need to be reinstalled.

In principle a minimal deployment is only for development/testing/PoC, whereas the enterprise deployment is a full production ready environment.

Need to ensure familiarity with the various deployment sizes in the reference architecture.

Details around each components availability etc is below.

Component Availability Notes
vRA Appliance DB automatically clustered (Manual promotion to master in synchronous mode), Active/active when configured behind load balancer
Model Manager Active/Passive - Configure two behind load balancer (Stop service on one node)
IaaS Web Active/Active - Configure behind load balancer
DEM Orchestrator Configure as active/active, but the secondary monitors the primary for failure and responds accordingly
DEM Worker Active/active. If a DEM-O detects failrue it will reschedule workflows on another worker
Proxy Agent Install both agents with the same name for high availability
vRO Configure behind load balancer. Exact configuration depends on load balancer type
MS SQL Windows failover clustering