VMware HCX improves VM mobility between vSphere clusters
VMware developed HCX to help customers move to VMware Cloud on AWS. HCX has since evolved as an offering and now enables VM and app migration between any two vSphere clusters.
VMware HCX enables VM and application mobility between vSphere clusters in different locations. Whether you migrate VMs and apps from an on-premises environment to a vSphere deployment in the cloud or between two different clouds, HCX simplifies the migration process and enables more flexible use of vSphere capabilities.
HCX, which is part of the VMware Cloud on AWS suite, was developed as a tool to extend your on-premises vSphere deployment into your AWS-based vSphere. HCX stands for Hybrid Cloud Extensions, and VMware built it to enable hybrid cloud mobility for applications inside vSphere VMs.
Now that most public cloud platforms -- as well as several service providers -- make vSphere available, you can use HCX as a separate product to help solve VM and application mobility problems.
HCX comes in two flavors: Advanced Edition and Enterprise Edition. Most vSphere cloud subscriptions with VMware or one of VMware's cloud partners include HCX automatically. HCX is also part of the VMware NSX DC Enterprise Plus product for an on-premises deployment.
What is HCX?
VMware built HCX to solve many of the problems inherent to VM mobility between physical locations. It enables the transfer of VM files from one site to another for easier movement of the VM itself. HCX also extends TCP/IP subnets between those sites so a VM can remain running and continue to deliver services during vMotion migration. It can optimize both storage and VM network traffic and provides deduplication and compression to minimize WAN bandwidth and cost.
VMware built HCX to solve many of the problems inherent to VM mobility between physical locations.
The Enterprise version of HCX includes additional capabilities. It automates the migration of VMs so you can migrate hundreds or thousands of VMs from one location to another at once. Migrate both powered-off VMs during an application outage or running VMs using vMotion. Enterprise HCX also has an option to enable disaster recovery from one location to another and then back again after an admin resolves the DR event.
Deploying HCX
You can deploy HCX as a set of virtual appliances. The HCX Hybrid Interconnect appliance handles connectivity between multiple locations and acts as the bridge between the local vSphere infrastructure and remote locations. It conducts storage and network traffic tasks between locations and encrypts all transfers between sites for privacy.
To optimize network traffic between locations, you should also deploy an HCX WAN Optimization appliance at each location. This appliance delivers deduplication and compression for better performance at a lower cost.
To enable live migration, you must deploy an HCX Network Extension Service appliance at each location. This appliance extends VM subnets between locations so VMs can retain their Ethernet MAC addresses and IP addresses as they move from site to site.
The current capabilities of HCX
Originally, VMware designed HCX to migrate VMs from on-premises data centers to VMware Cloud on AWS. Now, you have more options. You can use HCX to migrate VMs from on-premises environments to the VMware Cloud on other providers, such as IBM or OVH, or even between two different clouds.
By upgrading to HCX Enterprise, you gain even greater capabilities. HCX Enterprise provides a DR-to-the-cloud platform, which means you no longer must maintain your own DR data center. HCX Enterprise integrates with VMware Site Recovery Manager, a DR automation product. This integration enables regular and nondisruptive testing of your DR processes.
You can also use HCX Enterprise to migrate VMs that run on Hyper-V or Kernel-based VM into vSphere.
HCX's limitations
HCX cannot overcome issues with network bandwidth and latency. VMs can take time to replicate. Although bandwidth optimization helps, the size of a VM and the network speed between two locations ultimately decide how long replication takes. Large VMs replicated over slow networks might take weeks to replicate before a migration can occur.
The physical distance between locations dictates network latency between your locations. A high-latency link means your applications all live on one side of the link. For improved application performance, use HCX bulk migration to move all VMs that make up that application at the same time.
