For a successful cloud migration, creating (and later refining), a cloud migration strategy tailored to your organization’s goals, available resources, workloads, and priorities is an absolute necessity. So here, we’ll take a look at a simple list of the areas you’ll need to quantify and understand to build that strategy and then improve it as you move forward through the cloud migration process.
1. Recognize cloud migration options
You will need a clear understanding of the common scenarios that organizations encounter as part of a cloud migration effort. These options range from a relatively straightforward capture of an existing application workload environment to options for rewriting applications to wholly or partially depend on cloud-based resources or offerings to completely retiring an existing application and replacing it with an external Software as a Service (SaaS) offering.
There is also a last case that we won’t consider here – legacy custom applications that large organizations accreted over time that may be better left to “age in place” until it is time to replace.
“Lift and Shift”
“Lift and Shift” is the easiest of these cloud migration scenarios to recognize. Occasionally referred to as “rehosting,” in this case, an existing application workload (often including multiple dependent workloads) is captured from an existing set of physical or virtual servers and redeployed within a cloud or multi-cloud environment. As the name implies, the application is picked up and moved over.
“Extend to the cloud”
For mission-critical applications, one frequently used deployment scenario is to “Extend to the cloud.” In this case, an existing application workload is “extended” to use cloud service resources, most often cloud Infrastructure as a Service (IaaS). The business benefit gained is the scalability and cost-efficiency of the application, as cloud resources support “burst” modes when load increases precipitately on either a periodically predictable or unexpected event. Decreased costs result when applications no longer need infrastructure to support a maximum load at all times, while scale is delivered by using cloud infrastructure for higher loading events. Often the qualifying event for assessing an application as suitable for “extending to the cloud” is the need for high-speed transaction processing at the lowest layers of the application stack to support the existing architecture.
Keep in mind that this is a direct extension of the underlying application infrastructure, and doesn’t imply replacement with newer cloud-native technologies like serverless computing or containers and Kubernetes.
This cloud migration option is sometimes an “all in cloud” deployment and sometimes a “hybrid” deployment – that is, mixing cloud and on-premises resources. It differs from the “Extend to the Cloud” scenario by taking advantage of cloud resources beyond simple compute and storage. An existing application is rearchitected and partially re-written to take advantage of additional cloud service resources like AI and machine learning capabilities, cloud-deployed containers/Kubernetes, and other offerings while retaining a core of the existing application as is either on-site or in a public/private cloud deployment.
In this cloud migration scenario, an application workload is replaced by a new cloud-native application. It is rewritten and redeployed to make extensive use of Platform as a Service (PaaS) and serverless computing options, or a complete Kubernetes
Completely rewrite the application for cloud deployment. Containers, PaaS environments as well as the full gamut of available cloud services deliver the services of the legacy application, now optimized for cloud delivery, cost structures and capable of supporting the business with extensive new offerings as well.
“Replace with SaaS”
In many cases, once you’ve assessed an existing internal application, you may find that its functions are replaceable with a Software as a Service (SaaS) offerings. “Replace with SaaS” is an excellent scenario for applications that are not core to an organization’s success. HR, billing, salesforce management, and other functions are required for organizational success but don’t merit the attention of applications that can make or break your organization’s competitive stance and future.
2. Understanding your team’s capabilities
Your team’s skillset and training are a crucial part of setting your cloud migration strategy.
You’ll need to understand:
- Whether the team will be able to take on the implementation and management of migration and then on-going management of the migrated application without significant changes.
- Whether your team ready to learn and make the leap to master new skills sets for assessment, strategy setting, and deployment
- Even if your team already has all the right skills and capabilities, can resource loadings be adjusted (with the right trade-offs made) to undertake a significant new project (or project set)?
Once you have a clear understanding of your team’s capabilities, available time, and tools, you’ll be ready to make a call about whether outside resources are required. The next consideration is whether the scope of work needed justifies new resources, whether it is time to utilize a third-party to bridge skillset gaps or to do both to achieve a successful result.
3. Document goals and priorities
Now, let’s take a look at documenting your organization’s goals and priorities. We’ll do that in terms of three tasks.
Task 1 – The executive sponsor
When working on your goals and priorities, it’s imperative to have a strong executive sponsor (or set of sponsors) identified and driving the initiative. Trying to drive a cloud migration solely from IT operations, services, or development is a recipe for failure. Successful cloud migration initiatives need buy-in and context from key leaders that understand the priority and business context of the change. These leaders need to be aware of and support the resulting changes since digital transformation through cloud migration will result in substantial changes to the organization, roles, processes, and operation throughout. Even if IT operations or development are the leading-edge of the spear to get these changes underway, an entirely onboard executive sponsor or sponsor set needs to “be seen” as the principal focal point, and the owner for the desired business outcomes.
Task 2 – Initial priorities
Business units and executive sponsors may have either vaguely defined or specific priorities in mind for your organization’s cloud migration to achieve the business outcomes desired. Priorities can range from simple cost savings and scalability to improved customer experiences, to enabling entry into a new market area, or to addressing competitive threats with a revamped and restructured offering.
Task 3 – Specific, measurable goals
If cost savings is the critical measure, understand the goals from the organization and executive sponsors in terms of amount and scope. Similarly, if the goal is to radically enhance an existing application’s capabilities using cloud resources, an explicit set of goals is needed. Understanding the goals documented by this task will not deliver your final goals. Instead, this is about understanding business owner expectations and direction. The result should be an expectation set that you can then use to create and negotiate your initial, signed off, strategy.
4. First-level assessment of workloads for cloud migration
Now that you have an understanding of the organization’s goals and priorities, an understanding of what the options are for migration and the first-pass evaluation of your team’s capabilities and availability it’s time to assess those high-priority application workloads for migration to the cloud and the best deployment model to target for those workloads.
Assessing workloads is a complex task, to complete your initial cloud migration strategy, you won’t need a complete assessment for all your priority workloads, but you will need to understand the basics. For full details, see my page here on assessing workloads for cloud migration.
Here are the essential characteristics that you’ll need to evaluate your priority applications for:
Strategic value of the application
Is this application workload an element of your key differentiation or go-to-market? If the application isn’t of high strategic value, and unless the effort to transition is low, consider leaving it in place for now.
Application customization (if SaaS may replace commercial software)
Heavy requirements for application customization may be hard to achieve in many SaaS implementations. Make sure your organizational needs don’t prohibit a transition if SaaS offerings are under consideration.
Regulatory and privacy requirements
With repeated data-breaches and new personal privacy laws, careful assessment of meeting these requirements in a cloud environment can complicate your deployment and cause delays in implementation. Keep in mind that for security and privacy, the cloud is a “shared responsibility” environment. See excellent writeups and resources from Cloud Security Alliance.
Mission-critical applications with high-data throughput or transaction requirements are critical items to look for here. Compute performance requirements are relatively easily met and exceeded in the cloud, but storage I/O and transaction speeds can mandate hybrid usage.
Identify and bring in for essential decisions on trade-offs and implementation specifics. They are among your most critical customers for the project.
Many organizations are surprised to discover the number of internal applications and service dependencies for their applications. Make sure to audit this upfront, or you may misjudge the scope of your migration efforts.
SLAs and KPIs
Existing KPIs and SLAs will need to be maintained or exceeded – Know what you’ll need to achieve in the migration effort.
Application traffic and volume
Understand both the direction and scope of the traffic coming to your application. Both will affect cloud migration placement and scoping.
When cloud migration means “private cloud”
Applications with low variability, but high transaction or data access volume, are often best suited for migration to a private cloud. Variability enhances the benefits of a public cloud deployment.
Infrastructure cloud zones
Be sure to understand cloud vendor infrastructure zones. Especially when thinking about user experience and disaster recovery.
Infrastructure cloud costing and sizing
Any deployment to IaaS will require the selection and use of cloud instance and resource costs, speeds, and sizes. With major vendors offering 100’s of options, and options growing every day, a functional assessment of what you’ll need will help your organization to decide on the best cloud provider partner.
Cloud vendor lock-in – Application rewrites to use cloud-specific services, or extensive use of cloud storage, can “lock” your organization to a provider. Make sure you assess and understand what your migrated application will require.
Performance Assurance, Monitoring, Capacity Management, and Cost Management
Assessment of the tools you’ll need to cost-effectively assure the performance of migrated applications.
With this information, plus an assessment of your applications against the cloud migration options noted in item (1) above, you’ll be ready to move to the next step and formalize and initial strategy.
5. Formalize initial strategy
Now that you have a clear idea of the options available, a good understanding of your team’s capabilities, have documented the priorities and goals of the business, as well as possess a first-level level assessment of the cloud migration options best high priority application workloads, it’s time to create your initial plan with your stakeholders.
The job of planners and implementers here is to use what they’ve learned to assess difficulty and scope of work to migrate these priority applications to the appropriate cloud-deployment models and use this information to negotiate with business owners the initial strategy and deliverables. The outcomes of that work should include a set of specific priorities, metrics, and first-round scheduling, including:
- Selecting initial applications for cloud migration
- Profiles of the best cloud migration scenario for each application or application cluster, including critical considerations
- Initial project timelines and matrices of owners, tasks, stakeholders based on the migration type
- First-pass recommendations for cloud vendors and services to be used
- Plans for supporting the migrated application’s operation once migrated
What you’ll create at this stage is an initial strategy because no matter how thoroughly you’ve researched, documented, and assessed, it’s still true that “no plan of battle survives contact with the enemy.” In this case, applying this rule means that once your organization has worked through your initial deployments, you’ll have learned things about your infrastructure, applications, organizational dependencies as well as about cloud environment capabilities and offerings that you weren’t aware of at the start of the process.
This leads us directly to the next step – initial deployments and refining your cloud migration strategy as a result of what you will have learned.
6. and 7. Refining your strategy
Once you’ve initially deployed your first round of cloud migrated applications, what your organization has learned will be directly applicable to building out a full plan that meets your organization’s longer-term goals for digital transformation through cloud migration.
Use what you’ve learned and move forward.
Know Before You Go with Virtana Migrate
With Virtana Migrate, powered by Virtana Observe, you can accelerate and de-risk migrating existing workloads to the cloud. Plan smarter with our AI, machine learning, and data-driven analytics, combined with our deep understanding of your workloads running on premises, to understand your existing application workloads, infrastructure, and dependencies so you can intelligently prioritize workload transitions, and use process tracking and reporting to gain clear visibility. Request a free trial