Does this increasingly common scenario look familiar to you?
- Your organization has committed to using cloud services to run their computing workloads.
- New development will focus on containerized scalable microservices using cloud-native best practices.
- Your existing datacenters will be phased out.
What do you do with your extensive collection of existing applications running in those datacenters? Many of them are running on legacy operating systems and using legacy frameworks. Each application will need to be assessed, prepared, and migrated. Decisions will have to be made about what form of migration is appropriate for each workload:
- Lift and Shift
- move the application as is to one or more dedicated compute instances
- Pros:
- Quick
- Low effort
- Cons:
- resulting compute instances less efficient, more costly to run than containers
- you still have to manage and patch OS, infrastructure, security
- you must maintain the application and its dependencies as a monolith
- Lift and Optimize
- Lift and Shift, then optimize by extracting the workloads from their existing VMs into containers
- Pros:
- you no longer have to manage OS, infrastructure, security
- kubernetes provides management and scaling
- increased CPU utilization via containerization, fewer idle resources, lowering ongoing costs
- Cons:
- Moderate effort and time required
- Modernize and Migrate
- undergo development to modernize your application before moving it
- cloud-native, scalable design
- CI/CD
- Pros:
- maximize benefits of cloud
- optimized resources, lower ongoing costs
- accelerated development cycle
- Cons:
- High effort and time required
Once you have assessed your applications, odds are that:
- a few applications are appropriate to Lift and Shift
- most of your applications can benefit from the moderate effort of Lift and Optimize
- a few of your applications will be important enough to spend the up-front effort and expense of the Modernize and Migrate approach.
With medium-to-large collections of applications, this quickly adds up to a lot of moderate-effort projects, and you have a long, ardous road ahead of you.
What if you could make the Lift and Optimize process as quick and easy as the Lift and Shift process?
Google's Migrate for Anthos is designed to accelerate your migrations to the cloud by automating the containerization of existing applications and moving them directly from their existing VMs into scalable and manageable Kubernetes clusters, without touching the application code. This allows you to get many of the benefits of the cloud without expending the time and effort to completely modernize your application.
{% include image name="migration-paths.png" position="center" alt="migration paths" %}
figure: How Migrate for Anthos streamlines the migration process
For each application, you still assess the existing workload and plan the migration, but Migrate for Anthos steps in to automate the process.
As an organization with large scale needs, the benefits you get from using Migrate for Anthos to move those applications you would otherwise have had to Lift and Optimize yourself are:
- Save a significant amount of labor
- Dramatically decrease the time it takes you to:
- offload a large part of your workloads from your datacenter to the cloud
- realize the improvements in management, scalability, and cost enabled by containerization.
Let Migrate For Anthos launch your workloads into the cloud
Migrate for Anthos takes workloads that are currently on physical servers and VMs running on-premises or as compute instances in Google Cloud or other cloud providers, and converts and moves them into containers in Google Kubernetes Engine (GKE).
Google provides a Planning Best Practices document, which describes the types of applications that are best suited for Migrate for Anthos and also the workloads that are not a good fit due to specific performance, driver, or licensing needs.
//take the first step
You can start taking steps to accelerate your migrations now. The Getting Started with Migrate for Anthos documentation tells you what you need to know to get started with moving your applications from VMWare, AWS, Azure, or Google Compute Engine, including a tutorial on migrating a two-tier application from VMWare to the Google Kubernetes Engine (GKE).
//more to come
Keep your eyes peeled for further Arctiq blogs with more technical details about this exciting technology.
Photo by SpaceX on Unsplash