VMware Cloud on AWS
December 19, 2017
Get Started on Your AWS Hybrid Cloud Journey
VMware’s hybrid cloud service offering is referred to as VMware Cloud on AWS. VMware Cloud on AWS is a jointly engineered solution that takes the best of what VMware has to offer, from a software-defined virtualization platform, combined with AWS’s ability to provide a scalable, flexible and reliable cloud computing platform. The service is sold and supported by VMware.
For those who are familiar with provisioning a VMware cluster, they understand the time required in acquiring the servers, racking, cabling, installing the software and configuring it. What VMware has done is to allow the provisioning of a new cluster running on AWS within a couple of hours. The cluster is fully configured with vSphere (virtualization technology), vSAN (storage technology) and NSX (network and security virtualization technology) running on bare-metal hardware in one of the supported AWS regions.
In addition, customers get high bandwidth and low-latency access to native services within AWS to take advantage of other services available on the AWS cloud. It allows for operational consistency utilizing the same tools, policies and management capabilities that users may be already accustomed to. This means customers will not have to spend time learning new skills. Customers will have direct access to the vCenter server, although root ESXi access is not allowed. Customers will be able to scale to multiple SDDCs with each supporting up to 10 clusters.
Initial availability was announced at VMworld 2017 with a set of baseline services. However, VMware is committed to enabling more features on a regular cadence. Announced at AWS re:Invent 2017, the current offering is in two regions, US West (Oregon) and US East (N. Virginia). Moving forward, there will be support for all availability zones in supported regions. This expands capabilities for customers to run, move and protect mission-critical workloads at scale. The minimum size of a cluster is 4 hosts with the maximum being 32. Each host is configured with two 18 core processors for a total of 36 cores, 512 GiB memory and 15 TB raw NVMe flash storage with up to 25 Gbs throughput.
From a support perspective, VMware fully manages the SDDC within AWS. In the event of a host failure or problem, VMware will automatically deploy a new host. Data from the failed host will be migrated and/or rebuilt. The failed host will be removed from the cluster and the new host will be joined for you. If you want to expand your cluster, you can do so with the click of a button from the console or with an API call.
Recently VMware added Chat based support that allows customers to connect with a support representative immediately and open a case directly through the chat console.
Integration with 3rd party ISV solutions such as Veeam for backup, Ansible, Chef and Puppet for configuration management and CloudBees for continuous integration are just a few of the partner solutions that have been validated to work.
One important side note is that customers will receive two separate bills, one from AWS and one from VMware which includes support, software and hardware. It can be purchased through SPP (Subscription) or HPP (Hybrid) Purchasing Program credits. The service is billed as consumption based, with on-demand, hourly pricing on a per host basis. Customers can receive better pricing if they choose to go with a 1-year (up to 30% discount) or 3-year (up to 50% discount) upfront reservation. Data transfer and IP address charges will be added to the AWS bill.
Recent enhancements such as layer 2 stretch networking will allow customers to have one contiguous IP space across their on-premises environment and VMware Cloud on AWS. This will allow features such as live vMotion to move workloads back and forth without incurring any downtime.
With Hybrid Linked Mode, customers can link their on-premises vCenter server with the SDDC vCenter server even though they are not on the same version and not on the same SSO domain. This will allow customers to manage across environments from a single pane of glass. Keep in mind that the on-premises vCenter server needs to be running at least vSphere version 6.5d with layer 3 network connectivity to support hybrid linked mode.
In addition, the vRealize suite of tools around automation and operation can still be used for provisioning and management.
From left to right in the diagram below, in the blue box you have the customer owned and operated datacenter. AWS will manage the physical resources from a global infrastructure perspective. The blue box in the middle is the SDDC VPC that’s owned and managed by VMware. However, customers are responsible for managing their VMs running on top. The orange box represents the customer VPC and the two get connected during provisioning. Direct access to native services on AWS via private access is done using an elastic network interface (ENI). As long as the SDDC and the customer owned VPC are in the same availability zone, there are no data transfer charges. If they are in separate availability zones, then charges will occur.
Image courtesy of VMware
Common use cases for VMware Cloud on AWS include:
Dev/Test environments. This allows developers the capability to spin up and test their workloads without having to wait for hardware to be provisioned. Having the ability to deliver a consistent dev/test environment that integrates with common CI/CD toolsets and native AWS services enables new application development. If organizations are not able to innovate faster than their competition, they run the chance of becoming irrelevant and losing market share.
Having the ability to migrate workloads from your on-premises environment to the cloud at the click of a button brings about true workload portability. Nearly every company has an initiative to migrate some workloads to the cloud. VMware Cloud on AWS helps to remove those barriers by simplifying and speeding the provisioning of a cluster. Not having to spend the time to refactor applications as you move them from one environment to another is a big advantage.
Disaster Recovery as a Service (DRaaS). DRaaS with VMware Site Recovery simplifies disaster recovery operations. Typically, organizations would repurpose older equipment in a secondary site and hoped that it all worked when needed. With VMware Cloud on AWS, you can stand up a 4-host cluster running your critical services and then scale your cluster as needed. This scenario eliminates the need to support and pay for a secondary DC to support your DR efforts. Customers will be able to perform automated orchestrations and non-disruptive failover and failback testing.
Traditionally, organizations would need to do some sort of capacity planning and figure out what percent of utilization, typically 50-60%, at which they could run their clusters. This would allow some buffer room to handle a failed host or the addition of new workloads without having to procure more hardware each time. So, when you think about what having On-Demand Capacity really means, it brings the dynamic nature of the cloud to be able to elastically scale at a moment’s notice, whether it’s temporary, seasonal or planned.
Worth the Wait
VMware Cloud on AWS offers a lot of features that enterprise customers have been waiting on. It allows organizations to leverage the public cloud to enable a hybrid approach to modernizing their on-premises infrastructure. With VMware’s vSphere at its core, customers can maximize their existing investments while talking advantage of consistent operations delivered as a service from VMware.
For those who are interested in getting started on their journey, the team at Groupware can help assess your current VMware environment to determine what level of effort would be required to get you up and running. Our experienced and certified engineers will work with your team to design and architect a solution based on your desired use case as well as assist with deployment.
Daryl Harrington is Groupware Technology’s Practice Manager for Cloud Solutions.