Migrate workloads to the VMware Cloud on AWS by using VMware HCX
Created by Deepak Kumar (AWS), Derek Cox (AWS), and Himanshu Gupta (AWS)
Environment: Production | Source: On-premises VMware workloads | Target: VMware Cloud on AWS |
R Type: Relocate | Workload: All other workloads | Technologies: Migration; Hybrid cloud |
AWS services: VMware Cloud on AWS; Amazon VPC |
Summary
Notice: As of April 30, 2024, VMware Cloud on AWS is no longer resold by AWS or its channel partners. The service will continue to be available through Broadcom. We encourage you to reach out to your AWS representative for details. |
This pattern explains how you can use VMware Hybrid Cloud Extension (HCX) to migrate workloads from your on-premises VMware environment to VMware Cloud on AWS without changing the underlying platform. VMware HCX streamlines migration, helps rebalance workloads, helps protect data, and optimizes disaster recovery processes for both on-premises data centers and cloud servers. The pattern discusses the steps for installing, configuring, upgrading, and uninstalling HCX.
HCX supports the following:
Older versions of VMware vSphere – HCX helps you migrate virtual machines (VMs) from older versions of vSphere to VMware Cloud on AWS. The hosts are automatically updated and repaired to eliminate time-consuming updates in preparation for migration.
Bulk migrations – You can use HCX with a WAN optimization service to migrate a large number of VMs in one step without downtime, to expand your on-premises networks to the cloud.
Heterogeneous network environments – Your current network (such as vSphere, NSX, VXLAN, or NSX-T) determines the complexity of your migration. HCX extracts the fundamentals of your network application and extends your current network to the cloud without requiring any complicated procedures.
Slow network speeds – Migrations generally require connection speeds above 250 Mbps. HCX can migrate your workloads at much lower speeds, around 100 Mbps.
HCX supports three types of cloud migrations:
Hybridity (data center extension) – Extending an existing, on-premises VMware software-defined data center (SDDC) to AWS to provide footprint expansion, on-demand capacity, a testing/development environment, and virtual desktops.
Cloud evacuation (data center-wide infrastructure refresh) – Consolidating data centers and moving completely to the AWS Cloud (including handling data center co-location or end of lease).
Application-specific migration – Moving individual applications to the AWS Cloud to meet specific business needs.
You can use HCX to migrate workloads bidirectionally between your on-premises environment and VMware Cloud on AWS. HCX offers multiple ways to migrate your workloads between source and target locations:
HCX cold migration migrates VMs that are offline. This method is suitable for VMs that are powered off because it requires significant downtime.
HCX vMotion uses the VMware vMotion protocol to move VMs. HCX vMotion offers zero downtime migration but can migrate only one VM at a time.
HCX Bulk Migration uses VMware vSphere replication protocols to move VMs to the destination. You can migrate multiple VMs in parallel and schedule a switchover. The downtime is equivalent to a server reboot, and switchover for all VMs happen in parallel.
HCX Replication Assisted vMotion (RAV) is a combination of HCX bulk migration and HCX vMotion. It provides parallel migrations, scheduling, and zero downtime.
HCX OS Assisted Migration helps you migrate multiple VMs in bulk when you’re using multiple hypervisors and non-vSphere VMs on premises. HCX OS Assisted Migration is free when you use it to migrate from on premises to VMware Cloud on AWS, but requires additional licenses when you want to migrate between two on-premises environments or from on premises to other cloud providers.
Prerequisites and limitations
Prerequisites
A VMware account for access to the VMware console at vmware.com
. The following firewall ports are required for HCX.
Source
Destination
Port
HCX Manager and appliances IP on premises
HCX Manager and appliances IP on VMware Cloud on AWS
UDP 500, UDP 4500, and ICMP
HCX Manager and appliances IP on premises
connect.hcx.vmware.com | hybridity-depot.vmware.com
TCP 443
HCX Manager and appliances IP on premises
HCX cloud URL
TCP 443
If the on-premises network has internal firewalls, you will have to allow a few more ports locally within the data center. For a full list of port requirements for HCX, see the VMware HCX documentation
.
To configure HCX, you need the Domain Name System (DNS) IP, the vCenter fully qualified domain name (FQDN), the NTP server FQDN, the single sign-on (SSO) user, and similar information. Gather these details in advance to avoid any delays in the deployment.
Limitations
You can use the Network Extension appliance to extend a maximum of eight networks between the on-premises environment and VMware Cloud on AWS. For a full list of HCX service limits, see the VMware HCX documentation
Architecture
Source technology stack
On-premises VMware workloads
Target technology stack
VMware Cloud on AWS
Tools
Tools
VMware Cloud on AWS
is a service jointly designed by AWS and VMware to help you migrate and extend your on-premises VMware vSphere-based environments to the AWS Cloud. VMware Hybrid Cloud Extension (HCX)
is a VMware utility for migrating workloads from your on-premises VMware environment to VMware Cloud on AWS without changing the underlying platform.
Epics
Task | Description | Skills required |
---|---|---|
Enable HCX service in VMware Cloud on AWS |
| Cloud administrator, Systems administrator |
Generate the HCX activation key. |
| Cloud administrator, Systems administrator |
Add firewall rules for HCX on cloud SDDC. | After the HCX Manager is deployed, you need to configure firewall rules to enable communications between the on-premises environment and the SDDC. You need to create two firewall rules: one for inbound and the other for outbound communications.
| Cloud administrator, Systems administrator |
Install HCX Manager on premises. |
| Cloud administrator, Systems administrator |
Configure HCX Manager on premises. |
| Cloud administrator, Systems administrator |
Configure site pairing. | After you have configured HCX in the cloud and on premises, follow these steps to configure site pairing between them.
When site pairing is complete, the site pairing dashboard shows the on-premises and cloud SDDC connected. | Cloud administrator, Systems administrator |
Create a network profile. | A network profile is an abstraction of the Layer 3 components of a network. This profile is a prerequisite for creating a compute profile.
| Cloud administrator, Systems administrator |
Create a compute profile. | The compute profile consists of network, storage and compute details for HCX. HCX uses these settings when it creates HCX appliances during the creation of the service mesh.
| Cloud administrator, Systems administrator |
Create a service mesh. | The service mesh provides HCX service configuration for both the on-premises site and the cloud site. Creating a service mesh initiates the deployment of HCX interconnect virtual appliances on both sites. The interconnect service must be created on the source site.
Service mesh deployment will start. You can follow the progress in the Tasks tab for the service mesh. When deployment is complete, the status of all HCX services that you enabled for the service mesh is displayed. | Cloud administrator, Systems administrator |
Task | Description | Skills required |
---|---|---|
Create a network extension. | You can use HCX network extension capabilities to create a L2 network extension at the cloud SDDC HCX site and bridge the remote and source networks. This allows you to migrate servers from on-premises to VMware Cloud on AWS while retaining the same IP addresses.
When the network extension is complete, the system shows Extension complete. | Cloud administrator, Systems administrator |
Task | Description | Skills required |
---|---|---|
Configure replication. | To replicate VMs by using HCX:
| Cloud administrator, Systems administrator |
Task | Description | Skills required |
---|---|---|
Review recommendations and steps. | A large migration project can last from six to eight months, sometimes longer, and VMware periodically publishes HCX updates that consist of software fixes, security updates, and bug fixes. We recommend that you keep HCX and your appliances up to date to eliminate any security vulnerabilities and to take advantage of new functionality. Note: If your current HCX version is three versions behind the latest release or older, you cannot upgrade HCX and will have to redeploy it. An HCX upgrade consists of three steps:
The following stories discuss these steps in more detail. | Cloud administrator, Systems administrator |
Back up HCX Cloud Manager. | HCX Cloud Manager for VMware Cloud on AWS is managed by VMware, so you cannot take snapshots. To back up HCX Cloud Manager, you must download a backup from the HCX console and use this backup to restore the HCX configuration in case the upgrade fails or you have to roll back to a previous stage.
HCX service appliances such as HCX-IX, HCX-NE, and HCX-WO do not require individual backups. | Cloud administrator, Systems administrator |
Back up HCX Manager on premises. | You can back up HCX Manager on-premises in two ways: by taking a VM snapshot or by backing up the configuration file. To take a VM snapshot:
To back up the configuration file:
HCX service appliances such as HCX-IX, HCX-NE, and HCX-WO do not require individual backups. | Cloud administrator, Systems administrator |
Upgrade HCX Manager on premises and in the cloud. | You must upgrade HCX Manager on-premises first, and then upgrade HCX Cloud Manager. To upgrade HCX Manager on premises:
Follow the same steps to upgrade HCX Cloud Manager. | Cloud administrator, Systems administrator |
Upgrade service mesh appliances. | The service mesh is updated independently of HCX Manager at the source site. Service mesh appliances on the target site are updated automatically. To upgrade service mesh appliances at the source site:
| Cloud administrator, Systems administrator |
Task | Description | Skills required |
---|---|---|
Unextend network. | An earlier step explained how to use HCX network extension capabilities to create L2 network extensions and to keep existing IPs during migration from on premises to VMware cloud on AWS. When all the VMs from a particular VLAN have been moved to VMware Cloud on AWS, you must unextend the network between the on-premises site and the cloud SDDC, and make the network routable in the SDDC. We recommend that you remove the extended network as soon as all VMs are migrated from on premises to VMware Cloud on AWS to avoid latency.
| Cloud administrator, Systems administrator |
Route moved network in cloud SDDC. |
| Cloud administrator, Systems administrator |
Task | Description | Skills required |
---|---|---|
Check prerequisites. | In the case of a data center exit, we recommend that you uninstall HCX and remove its components at the end of your migration project. However, if you still retain an on-premises footprint, you might want to keep HCX running. Before you uninstall HCX, make sure that:
| Cloud administrator, Systems administrator |
Uninstall HCX on premises. |
| Cloud administrator, Systems administrator |
Deregister HCX plugin from on-premises vCenter server. |
When all extensions have been removed, the HCX plugin will disappear from the vSphere Web Client. | Cloud administrator, Systems administrator |
Uninstall HCX in the cloud. | To remove the HCX service mesh and site pairing in the cloud, repeat the steps described earlier in Uninstall HCX on premises. In VMware Cloud on AWS, HCX Manager is managed by VMware. You cannot delete it from vCenter, but you can undeploy it from the VMC management interface. To undeploy HCX Manager:
| Cloud administrator, Systems administrator |
Troubleshooting
Issue | Solution |
---|---|
You’re unable to select the servers to migrate when you configure HCX bulk migration. | Cause: Migration for these servers were canceled, but the HCX database wasn’t updated during the cleanup. HCX views database migration as still being in progress, so it has locked the status at "Switchover in-progress." Solution: Reach out to the VMware support team to clean up the HCX database. |
Switchover fails but works with the Force Power Off option. | Cause: The version of VMware Tools didn’t meet the prerequisites for HCX bulk migration, so HCX could not shut down the source VM. Solution: Update the VMware tool to the recommended version for your migration type. |
HCX site pairing appliance upgrade fails with the error "Operation not allowed for ongoing bulk migration" while migration is in progress. | Cause: The HCX database didn't update after the switchover. Solution: Make sure there are no ongoing migrations. Choose Force upgrade when you upgrade the site pairing appliance. |
Cutover fails with error "Low resource availability." | Cause: Low storage on the host VM. Solution: Check storage and compute resources before migration. |
Related resources
References
VMware Cloud on AWS overview and operating model (AWS Prescriptive Guidance)
Migrate VMware SDDC to VMware Cloud on AWS using VMware HCX (AWS Prescriptive Guidance)
VMware HCX in the VMware Cloud on AWS
(VMware documentation) HCX HCX release notes
(VMware documentation) SDDC Deployment and Best Practices Guide on AWS (AWS whitepaper)
Tools
VMware Cloud on AWS Automation using PowerCLI
(VMware Cloud Tech Zone)
Partners
Videos
VMware Cloud on AWS
(YouTube video)