Many enterprises face the daunting prospect of migrating hundreds—or thousands—of applications to the cloud. It's no wonder that a lot of them are tempted to take a piecemeal, ad hoc approach.
But by doing so, they may miss the opportunity to optimize and modernize their applications and make platform decisions that support their organizations' strategic goals. Without proper planning and analysis, you also risk encountering surprise issues that may create ripple effects down the road.
A factory approach to cloud migrations makes large-scale digital transformations manageable by applying the principles of industrialization. This approach accelerates cloud migration projects and reduces complexity by:
- Identifying key migration patterns
- Grouping applications into those patterns
- Using predefined, repeatable processes to standardize and scale migrations
During the discovery phase, a company with 2,000 applications may find that they fit into 20 migration patterns or "treatments." Once it has identified those treatments, the company can use a cloud migration factory approach to perform migrations at scale by following its standardized treatments.
With a cloud migration factory, each application moves through an assembly line of specific steps on its way to the cloud: readiness assessment; preparation for the cloud; identification of the best treatment and target platform in the cloud; remediation, testing, and validation; and deployment.
This approach lets you perform large migrations at high throughput, with maximum efficiency, and without introducing risk into your environment. Cloud migration factories can speed application migrations by up to 80%.
Properly implemented, this approach virtually eliminates the possibility of failure. Here's how it works.
Take a structured approach to app migration
Before beginning a cloud migration, IT executives and business leaders should take a step back and craft a strategic plan that encompasses business challenges and goals, skilled resource requirements, a thorough assessment of their current applications, and the deployment of automated processes and specialized tools.
Once you have established an overall strategic plan, you're ready to begin the migration project. Major steps include:
- Creating a readiness road map, building the business case and assessing the total cost of ownership (TCO) of the company’s entire applications portfolio.
- Assessing the application estate, making key decisions about the migration path for different types of applications. You should prioritize applications based on their value to the business: Are they customer-facing or revenue-generating? Must they be continually available or handle rapid fluctuations in demand? Some applications need to stay on premises for regulatory, security, and business process reasons. Some may not be worth the cost of rewriting because they are near the end of their lifespan.
- Identifying key patterns and selecting the correct migration treatment for each application based on the application requirements and TCO model.
- Deciding where each application will land: public cloud infrastructure, on-premises private cloud, hosted private cloud, software as a service (SaaS), platform as a service (PaaS), containers, cloud databases, serverless environments, etc.
- Performing an initial set of migration sprints.
- Creating a secondary road map that addresses applications not included in the initial rollout.
- Making a detailed migration plan, including deep discovery to keep migration activities on schedule.
- Making "day two" operations a priority. Once an app is migrated, make sure it's running and the team knows how to operate it in the target environment.
Incorporate automation and agile development
Your migration factory should automate as many steps as possible, including remediation planning, the assignment of remediation tasks to appropriate experts, testing, and deployments.
It’s important to determine the disposition of applications early in the cycle, and automation can assist in that process as well.
The simplest way to determine the appropriate treatment for an application is to run it through a predefined rules engine. Today's more advanced, modern models use machine learning to evolve and become more accurate.
These tools compare the original prediction for application treatment to the actual treatment you used. No single automation tool does it all, however, so you'll still need experts to orchestrate the work and the tools.
It's also key to use an agile approach that minimizes your operating investment. With an agile model, you do the work incrementally in two-week sprints, each sprint building on the work of the previous one as you make continuous improvements and adjustments.
A one- or two-step approach?
When deciding how to migrate an application, think about whether you should take a one-step or two-step approach. In the one-step journey, you modernize the application before migration and move to the cloud in one fell swoop. In the two-step method, you move the application to the cloud first, then modernize it using cloud-native tools and methodologies.
The two-step approach is the most common choice because it delivers benefits faster, provides more cost transparency, and reduces risk because it has minimal impact on the application during migration. It also eases future application modernization and improves security by moving applications to a micro-segmented network model.
Most companies realize significant cost savings once the application has been modernized, but the two-step approach takes longer and pushes savings further into the future.
Companies that are more risk-tolerant and focused on reducing TCO in the near term may lean toward the one-step approach. Consider this approach if you need to reduce third-party licensing costs, exit end-of-life platforms, or improve the user experience by using new user interface (UI) technologies.
But when you take the one-step approach you risk biting off more than your organization can chew, plowing ahead before you've resolved important issues such as organizational silos or lack of skill sets needed to do the work.
There is no right answer for choosing between a one-step and two-step approach; it depends on the situation for each application, and many companies use both methods. The most important thing is to carefully consider which approach best aligns with your business objectives. This will help ensure the success of your migration efforts.
Keep retooling your factory
It's important to take a holistic view of cloud app migration. This is not a "one-and-done" activity, but a continuous, iterative process. Review and revise your plan as needed, incorporating lessons you learn along the way, continuously improving it to optimize performance, and adjusting it to accommodate your evolving business priorities.
Finally, don’t forget to include security in every step of the process and to prioritize training and transfer skills to your DevSecOps team.
Keep learning
Get up to speed on digital transformation with TechBeacon's Guide.
Download this free IDC white paper, "Enabling End-to-End Digital Transformation".
See IDC's Futurescape: Top 10 predictions for digital transformation.
How important is digital transformation to your org? Take our survey and find out how you stand next to the competition.
Thinking of making a change? TechBeacon's Careers Topic Center provides expert advice to prepare you your next career move.
