Wen Chi Li – Group CTO and Head of Global Delivery, Cloud Kinetics
Puneet Arora – Cloud Solution Architect, VMware
Sawab Ahmed – Partner Solutions Architect, AWS
Zhihao Pang – Solutions Architect, Cloud Kinetics
- “VMC on AWS – Cloud Migration Strategy: What to look out for” by VMware
- “Modernisation with VMC and AWS Native services” by AWS
- “Use case and scenario for VMC on AWS” by Cloud Kinetics
- Q&A session with the presenters
Key excerpts from the webinar:
Customers are urgently moving to cloud to optimize their applications, reduce costs, and unlock innovation.
- 47% of technology decision makers believe they are behind schedule in their cloud journey due to migration complexity.
- 7.4 year to refactor and migrate 100 applications to the cloud.
- Modern apps are deployed in the cloud and on prem. 49% apps in data centers will be modern in 3 years.
With VMware Cloud on AWS:
- Leverage the benefits of on-premises environments and public cloud environments.
- Set up and entire software-defined data center in the cloud in under 2 hours and scale capacity within minutes.
- Faster ramp up time with the management interface of the vCenter having the same look and feel and operates the same in the cloud as it does on-premises.
Free PoC – Try before you Buy!
- Validate your VMware VMs on VMC on AWS with a clear billing TCO monthly forecast, paying-what-you-use-only
- Validate VMC on AWS use cases such as FSx / S3 integrations with clear monthly forecast, paying-what-you-use-only
- Validate your VMware VMs on VMC on AWS multi-tenancy environment SDDC / CDS, without paying for any ESXi Nodes
Q&A on VMC on AWS Migration:
1. Is there a minimum node required when deploying VMware Cloud on AWS
with Cloud Kinetics?
Traditionally, if you’re running a single tenancy on VMware Cloud on AWS, the minimum number of nodes you can have per cluster is two. CK allows you to run a multi-tenancy model, i.e. you don’t have to make the investment of the minimum number of nodes actually required.
We have made the investment for you. You can run any available VMs that you will actually require. For example, it can range from one VM to maybe a few hundred VMs. It really depends on what your use case is. You only need to pay for what you use. In the event that you only require one VM to run on VMware Cloud on AWS, it is certainly feasible because of this multi-tenancy model that you run with CK.
2. What is the approach with VMware HCX, to AWS Snowball and AWS Outposts?
VMC on AWS Outposts is basically our service offering which is built upon the same architecture as VMware Cloud on AWS. The only difference is, this now resides inside your data center itself. So, VMC on AWS is predominantly built for customers who, because of data sovereignty or because of local cloud requirements, want to host the workloads on VMC on
AWS, but do not want to take it away from their on-premises. This is where AWS Outposts comes into the picture. Some of the migration methodologies in VMware Cloud in AWS remain the same. We can still use HCX, which is optimized to migrate workloads even for low-bandwidth, high-latency links. There’s an optimisation engine built inside HCX which allows you to migrate even if you don’t have a healthy pipe to do the migration. That’s the benefit.
Besides that, HCX not only offers you an online migration methodology, it also offers you an offline migration methodology. You can actually shut down the VMs. You can export it in a standard way, then you can print them one by one, and so on and so forth. There are a couple of options. Snowball currently is not a support option, but that’s the roadmap. There are enterprises who are really looking to migrate out their data and they want to leverage Snowball as a capability to take out the data and move it over to the edge location where they want to place names in individual outposts. We can definitely explore that as a possible option but down the line.
This is how the connectivity happens over on our VMware Cloud on AWS Outpost service. It is not available in all VMC regions for now, just as a disclaimer. This is how the connectivity works from your AWS output service to the external world. For example, if you are talking from an output service, you’ll see here that there will be a VMware-managed AWS account where your output service will lie, which is in one availability zone. There could be another availability zone, which is a customer-connected VPC, for example, where you will have an AWS-native VPC. And then, you will have a customer on-premises account where your AWS outposts rack will be located. You have all this connectivity set up here where you will have a service link, which will be accessed from your AWS outpost service to your AWS outpost rack. The connectivity from your outpost rack will be different. This goes onto an ENI connectivity then onto your connected VPC. AWS Snowball is different. It’s architecture. It does not use the same architecture as AWS Outposts.
3. What is Cloud Kinetics’ plan to better support customers’ cloud journey?
Cloud Kinetics is a cloud-first company. With today’s choice of cloud solutions and deployment options, you want to navigate through all these options that require new cloud skills and infrastructure strategies, which most organizations lack. So, we partner with our customers to help shorten this migration path and achieve whatever business objectives the customer has. We will address uncertainties by reducing the need for trial-and-error and side-stepping common challenges. In addition to any cloud advisory, organizations can take advantage of the growing number of programs that CK designs to get them through any cloud implementation. They may see a significant advancement in these commercial programs and tools, and our customers will benefit from them. We also bring to the market, a multi-tenant version of the VMware Cloud so you do not have to make that significant investment in terms of the multiple clusters, no matter the number of VMs you want to have.
4. How long will it take for the migration of 5 to 8 VMs to be done?
VMC migration happens predominantly in a couple of ways. We have methods of doing it online as well as offline. And, more often than not, you would want to go with an online migration because you don’t really want any downtime while migrating. That HCX is the secret sauce that allows you to do that. The time it takes to migrate the VM is basically a factor of a few things: the number of VMs and the capacity of the VM you’re trying to migrate. You’re doing an online flat-file copy. Given the fact that we have an optimisation, and we have network deduplication methodologies to make the migration process faster, it will not take much time. Bandwidth is also another important factor. The link between your on-premises data center to the VMware Cloud and the geographical distance to your cloud tenant versus your on-premises location are factors that can decide this. You can actually see in the HCX migration interface, when you initiate the migration, that it can estimate all these factors and it can give you an estimate of how long migration would take. We have customers who had a good link and pre-prepared the migration bill. We have managed to do hundreds of migrations over weeks. Five to eight VMs can be done in a matter of hours, or days at the max, provided we have all the pre-planning and prerequisites taken care of.
5. Will the VMC on AWS solution help in migrating applications with persistent storage?
Yes. With the very recent release of VMC, the m16 version released about a month back, we now support Tanzu services on VMC. So, if you’re running a TKGI infrastructure, which is on-premises running on Tanzu Kubernetes grid, and you would want to host a similar environment on VMware Cloud on AWS using our DNS services, it is supported because we now have tons of services available to us. That’s one part of the solution. Tanzu, like any Kubernetes platform, is a stateless architecture. So, generally, that DR doesn’t happen the same way you would do a DR for a standard virtual machine. If you would want to retain the state of your persistent storage, which is sort of a stateful component of the modernized application, we can definitely do that. We have methods of backing up and restoring on TKG. We can do it using some of our DR services like VCDR or VMware site recovery. All that is possible today.