Reference architecture for setting up multi-site HCX with VMware Cloud on AWS is now available, helping you to visualize what needs to be deployed in your SDDC so you can easily migrate workloads to the cloud. This article explains the main points of the reference architecture.
Are you planning to extend multiple on-premises data centers into a VMware Cloud on AWS SDDC using VMware HCX? Reference architecture is now available to help you on the journey.
What is VMware HCX?
VMware HCX is a mobility and data centre migration tool that enables the rapid migration of workloads to the cloud without the traditional requirements of legacy hardware upgrades, installing additional hardware, or the installation of dedicated circuits from an ISP. It provides L2 extension, bulk migration and vMotion (note that, for vMotion, distributed switch is required on networks for source VMs).
Using the reference architecture
The connectivity shown in this reference architecture is set up during the deployment of HCX; and is in addition to that which would initially allow the on-prem locations to connect to and use services in the VMware Cloud on AWS SDDC. (See the baseline architecture “Getting started with VMware Cloud on AWS”.
In this HCX reference architecture, we concentrate on what you need to deploy on-prem and what gets deployed into your SDDC to enable replication/migration of workloads and the extension of networks. Within the SDDC, we only need to deploy one HCX manager, but we need to deploy the Interconnect, Optimization and Extension appliances (the “fleet”) for each on-prem site you want to connect to the SDDC.
Although we show the Extension and Interconnect IPSec tunnels between the devices which they connect, the physical path between those devices will be the same (red) Internet connection – you’ll need 100Mb/s of Internet access at your on-prem locations for good performance. As is often the case with networking, more is better and, even if you have high bandwidth connectivity, if it’s already “full” before you add HCX, you should probably read that to mean “an additional 100Mb/s of spare throughput on top of what we have.”
While this reference architecture shows two on-prem locations connecting to an SDDC, HCX is capable of creating a multi-site service mesh where different sites can be used for different roles. Watch out for an upcoming blog post explaining more on that, or check out the HCX documentation.