This content is part of the Essential Guide: VMware virtual recovery and backup best practices and tools

VMware Cloud on AWS shows potential despite limitations

Powerful specs of VMware Cloud on AWS point toward better bursting to the cloud for disaster recovery, but future success relies on ease and flexibility.

VMware's new partnership to bring its infrastructure to Amazon Web Services arrives with some initial limitations for workload movement but has potential for bursting to the cloud for disaster recovery. VMware Cloud on AWS is likely to benefit commercial and enterprise users the most because of their needs for security and ease of use.

Powerful specs for large buyers

As of its release, customers are limited to a single host configuration. You're looking at dual CPUs with 36 cores -- 72 with hyper-threading -- 512 GB of RAM and 10.7 TB of flash storage. The service also requires a minimum entry point of four hosts. This is a lot of horsepower, and it's clearly targeted at larger customers. Do you need an environment this powerful?

Complications with bursting to the cloud for DR

Two use cases for the service initially come to mind: bursting to the cloud and disaster recovery (DR). The ability to transport VMs from an on-premises environment to one in the cloud could cut down critical infrastructure needs from weeks or months to hours. This is the tagline, and it's true for the most part.

While you can provision an environment quickly on VMware Cloud on AWS, you still have to configure it. You have to deal with not only the external interfaces to clients but also with the connections coming back to the on-premises data center. The network is part of that configuration, and administrators must still complete the traditional VMware farm configuration steps in this new environment, including setting up resource pools, high availability and Distributed Resource Scheduler. While these steps aren't intensive, they are time-consuming and necessary for every new deployment.

While you can provision an environment quickly on VMware Cloud on AWS, you still have to configure it.

This can add a lot of additional work to what might already be an overworked staff. Administrators can automate repetitive tasks, but it still takes effort to set up that automation. If the burst need isn't always the same, you will need administrators to adjust the process each time.

The burst aspect isn't as clean or as easy as it might appear, and while it still has value, you need to take these aspects into account.

The costs of contingency planning

You're often paying for capacity or resources with DR that you hope you never need. While some redundant equipment and environments might be available to handle additional tasks, they are often idle and collecting dust.

In cloud environments, needing to be on standby doesn't reduce the cost of the environment. You are billed for the reserved capacity, not what you're actually using. This can lead to some sticker shock from the higher price points of VMware Cloud on AWS, especially when it's on standby. DR to the cloud might not be ideal unless you can do it on demand, which is now possible given how quickly you can spin up environments in AWS. This would only be ideal in certain circumstances, such as natural disasters approaching slowly enough to enable a switch to the cloud environment.

VMware Cloud on AWS as a platform

It's possible to launch a new business unit or core application -- such as a new healthcare record system for a hospital -- using VMware Cloud on AWS. Rather than buy the infrastructure for in-house deployment, you could spin up the infrastructure on AWS and go with an Opex model instead of a Capex model.

While that sounds ideal for larger-scale applications, many larger applications now have SaaS offerings. It can be hard to compete with the SaaS offerings as it might not be as cost-effective based on licensing, maintenance and support.

Relevancy of VMware Cloud on AWS

Is VMware Cloud on AWS relevant, especially considering the existing configuration and costs? While it isn't a niche-use item, it won't apply to everyone today. Currently, data center moves or longer-term bursting work better for those who need to move workloads around.

The key to the success of VMware Cloud on AWS in the future will be its use of automation and bandwidth. If it becomes possible to burst workloads into isolated infrastructures quickly and easily with tools such as vRealize Automation, this could be a huge benefit to many commercial and enterprise users who need those security and isolation aspects but with the ease of automation. It's all going to come down to ease and flexibility -- both things AWS has done very well. It's up to VMware to follow that example as it moves forward with VMware Cloud on AWS.

Next Steps

The limits of cloud bursting

Use the cloud for disaster recovery

Use automation to move cloud workloads

Dig Deeper on Cloud computing architecture