Introduction
Did you already hear about Workload mobility? Datacenter relocation with zero downtime? Do you have a company GoToCloud strategy?
Believe it or not, but moving a workload from a datacenter to another one without any service disruption is now something that became true thanks to HCX!
When you think 15 years earlier the revolution of moving a virtual machine (or workload) from a physical server to another physical server was handled by vMotion. Today we are pushing the magic to the datacenter move with HCX!
We are now in the era of multi-cloud. We have workloads On-Premise or in a traditional service provider. In addition, we now may have native applications running in Azure, Google or Amazon. Many companies have a multi-cloud strategy but they have no idea how to move in a public cloud without taking risks and mitigating the risks without huge costs associated with this approach.
The major difficulty is to re-platform, as this will drive a lot of team efforts and so a cost. The project time will be challenging and the rollback not easy.
HCX will remove all these rework costs and will reduce the migration project time by >70% and reduce the cost of migration by 53%. You will be able to migrate to any cloud without downtime by just doing lift and shift of the VMWare workloads to a VMware cloud provider compatible offer. The staff required to manage the migration will be reduced by approx 59%.
This as well offers you the flexibility to exit any public cloud provider at any time to avoid vendor lock-in as you could move your workload from anywhere to anywhere.
Unexpected downtime for your workloads will be improved by 83% and will cost from an infrastructure prospective around 39%. (all these figures are based on IDC report).
This solution will provide a number of capabilities depending on what you need and what you need to accomplish.
Service Overview & Licensing
The official HCX guide is HERE
HCX is relying on different services and is available in 2 licensed versions :
- HCX Advanced
- HCX Enterprise
SERVICE DETAILS
As the limitations and configuration maximum are moving fast you may require to double-check on the official documentation HERE
if there are any updates on the numbers from this article.
-
Hybrid Interconnect (HCX Advanced & Enterprise)
The HCX Service provides resilient access over the Internet and private lines to the target site while providing strong encryption, traffic engineering, and extending the datacenter. This service simplifies the secure pairing of sites and the management of HCX components. The service is deployed as an HCX virtual appliance (HCX-WAN-IX).
You can interconnect up to 25 destinations per source HCX manager.
This appliance IX appliance will be seen as VMware Mobility Platform in the vCenter.
We will talk more in details in the Migration Article that I will release later ;o)
-
WAN Optimization (HCX Advanced & Enterprise)
The WAN Optimization service improves performance characteristics of the private lines or Internet paths by leveraging WAN Optimization techniques like data deduplication and line-conditioning. The service is deployed as an HCX virtual appliance (HCX-WAN-OPT).
If you have low bandwidth or if your express route is already heavily used for production flows, you could optimize the migration traffic by enabling this option.
-
Cross-cloud vMotion Migration (HCX Advanced & Enterprise)
The vMotion Migration service provides zero-downtime, bi-directional Virtual Machine mobility. The service is deployed as an embedded function on the HCX-WAN-IX virtual appliance.
This implies having the same physical CPU funder on each site (For AVS the target physical servers are built on INTEL pCPU). IF your pCPUs are not eligible for vMotion you will have to use the Bulk Migration Option which implies a short downtime when the switchover occurs on the target Datacenter.
The migration will occur in sequential mode 1 VM at a time. The minimum VM hardware will require version 9.
-
Bulk Migration (HCX Advanced & Enterprise)
The HCX Bulk Migration service provides low-downtime, bi-directional, Virtual Machine mobility using VMware vSphere Replication technology. The service is deployed as an embedded function on the HCX-WAN-IX virtual appliance.
Bulk migrations are limited to 200 per service mesh. Minimum VM hardware needs to be at least in version 7.
-
Replication Assisted vMotion Migration (HCX Enterprise Only)
VMware HCX Replication-Assisted vMotion (RAV), combines advantages from VMware HCX Bulk Migration (parallel operations, resiliency, and scheduling) with VMware HCX vMotion (zero-downtime virtual machine state migration).
This methodology allows you to perform up to 200 concurrent VM migration. The minimal virtual hardware required is version 9 at the VM level.
-
Network Extension (HCX Advanced & Enterprise)
High throughput Network Extension service with integrated Proximity Routing unlocks seamless mobility and simple disaster recovery plans across sites with IP/MAC address retention. The service is deployed as an HCX virtual appliance (HCX-NET-EXT).
This is a key feature in the migration simplification as the Virtual Machine will keep the same IP from the Source Datacenter and the Target Datacenter.
The limits are 8 network extension per HCX virtual appliance (HCX-NET-EXT). You can have up to 10 HCX virtual appliances.
-
Disaster Recovery (HCX Advanced & Enterprise)
The Disaster Recovery service is a business continuity solution used to protect workloads from app-level to site-wide disasters. The service is deployed as an embedded function on the HCX-WAN-IX appliance or in a scale-out fashion (HCX-DR).
VM protection is limited to 500 VMs per HCX manager while I’m writing this article.
-
SRM Integration (HCX Enterprise Only)
Provides the ability to configure DR from VMware Site Recovery Manager. HCX can be used as a SRM extension for the replication of VMs.
SRM could be a great approach for a massive big bang migration approach as SRM can automate the restart order and re-IP of the VMs at the destination site.
-
OS Assisted Migration – Also called OSAM (HCX Enterprise Only)
The OS Assisted Migration service can be used to bring OpenStack workloads into a vSphere environment. To use this service click the Interconnect Tab on the home page and then click on the ‘Sentinel Management’ Tab.
This migration approach will be subject to a dedicated article. You could as well migrate several types of non-source VMware workloads.
See OS Assisted Migration Limitations if you can’t wait until I publish the dedicated article.
Appliances form factor
Depending on the services you want to enable on HCX, you need to understand that these services will rely on virtual appliances automatically deployed by the HCX manager based on the compute profile and service mesh setup (This will be the next article I will publish).
To help you understand the capacity requirements to deploy such architecture find below some figures.
The official documentation is HERE.
Appliance | # vCPU | Memory (GB) | Disk Space/IOPS |
---|---|---|---|
HCX Manager | 4 | 12 | 60 GB |
HCX-IX (Hybrid Interconnect & vSphere Replication) | 8 | 3 | 2 GB |
HCX-NE (Network Extention) | 8 | 3 | 2 GB |
HCX-WAN-OPT (Wan Optimization) | 8 | 14 | 100 GB / 5000 IOPS |
HCX-SGW (source only – OSAM) | 8 | 3 | 10 GB |
HCX-SDR (destination only – OSAM) | 8 | 3 | 10 GB |