It might seem that there is a bewildering confusion in Dwight D. Eisenhower’s famous quote: “I have always found that plans are useless, but planning is indispensable.” How could plans be useless and planning is indispensable? Yet, this was one of our main lessons learned from our cloud migration initiative.
In February 2019, the National Council on Aging (NCOA) achieved a key milestone in its IT strategy. We completed 95% migration of our IT assets to the Amazon Web Services (AWS) cloud.
The migration effort included our financial system and our digital products. We also made major progress toward establishing our Digital Workplace, which included migrations to Office365 with OneDrive, SharePoint, and the relaunch of our Intranet.
This was a coordinated effort as part of a larger cloud migration strategy that took almost a year to complete. There were many lessons learned from this complex effort, but the not-so-surprising and clear lesson is that planning is everything—and that migration to the cloud is more of a strategic than a tactical move.
Migrating from an on-premise IT infrastructure to the cloud can be perceived as a tactical move toward getting rid of complex, messy, time-consuming, undifferentiating infrastructure operations. Cloud hosting also offers immediate scalability and elasticity in providing the right computing power required by the organization and its customers.
However, with recent advancements in cloud capabilities and offerings, combined with the relentless shifts in the business environment that force businesses to continuously innovate their products and optimize their operations, cloud migration is amounting to a strategic move.
Many public cloud offerings include capabilities such as analytics, compute, augmented and virtual reality, blockchain, storage, security, AI/ML, and robotics-enabling capabilities. These capacities offer organizations a path to quickly innovate and experiment, which ultimately enhances their competitiveness. Between the powerful cloud technology offerings and organizational needs for digital transformation, cloud migration is clearly a strategic move.
Our migration to the cloud was driven by our business strategy and digital transformation initiative. Serving tens of millions of customers with a wide variety of health and financial-related products requires scalable IT infrastructure with elasticity and out-of-the-box tools that can enable immediate and inexpensive innovation and experimentation.
Our cloud migration goals also included achieving operational excellence, optimizing the utilization of our IT assets, and mitigating risks -- especially when it comes to security, Disaster Recovery (DR), business continuity, and removing personnel single points of failure. The migration to the cloud also offered the opportunity to generate a better value for hosting our IT assets.
Throughout our cloud migration effort, we followed the classical cloud migration methodologies and plans that included the following phases:
Below is a detailed list of our plans and how we implemented them:
- We defined the business drivers behind the cloud migration and what business, IT outcomes and values we expected. We understood the business and IT operations added-values and business enablement offered by industry-standard cloud hosting environment.
- We assessed our existing IT portfolio and inventoried all assets. We assessed our digital products, ERP, CRM, and our internal productivity and collaboration tools. We were then able to write the requirements for the migration effort. This defined which assets must be migrated, where to, along with their technical capabilities and infrastructure needs.
- We developed a request for proposal (RFP) based on the assessment and requirements. We invited 10 vendors who were partners of AWS, IBM, Google, and one private hosting vendor. Through a detailed RFP process, we selected a vendor that met our requirements and can be a future partner beyond the migration effort.
- We developed migration plans, including which environments and products to migrate, when, and in what sequence. We also developed testing and verification plans.
- We migrated in phases, based on our plan. We focused on the less risky and simpler to migrate products first. Logically, we migrated the QA environment in the first phase.
- We developed simple operational KPIs, SLAs, and analytics to monitor the performance of the cloud infrastructure and the performance of the MSA partner. Over time, our monitoring dashboard will include reviewing the business value of the migration to the cloud and not just the operations.
After selecting the AWS cloud infrastructure and vendor, we decided to apply the traditional AWS migration methodology -- the so-called lift and shift. When we completed the migration effort, we conducted a post mortem evaluation and lessons learned exercise. The following are the high-level lessons learned from our migration effort:
- Detailed planning is a critical success factor. This was the most important lesson. No matter how prepared we thought we were, when we migrated the more complex IT assets, we discovered that the planning and time allocated to the migration effort were not enough. We recommend investing more time in migration planning, especially in QA/QC planning and in test scripts development.
- Lift and shift proved to be a good migration strategy. While we followed the lift and shift AWS methodology, we still did minor tinkering by moving our MySQL databases into AWS Aurora RDS to provide better high-availability and DR services. In retrospect, this was the right decision. However, it requires an additional data validation effort
- Dissimilar production and QA/QC environments proved problematic. As an organization with limited resources, we had to make tradeoffs. One tradeoff was to have a QA environment that is not identical to production. This proved challenging and problematic when many last-minute production issues could have derailed the cutover and project timeline. In retrospect, we should have created an identical QA environment from a data content perspective with a stripped-down topology and specifications.
- Migration in phases was beneficial. We started with the less complex and risky system, which proved beneficial and created some early lessons learned. However, this also created a higher self-confidence and false impression that the more complex systems would be just as easy to migrate. Migrating in phases is the right decision, but make sure to be just as rigorous in planning and testing efforts.
- Staff training is not enough. We trained our staff on the AWS environment, but training is not enough to guarantee success during migration. More training would have helped, but additional temporary internal and experienced resources might have improved the migration effort.
- Strategic partnership with the migration vendor is critical. We selected our migration partner to not only provide AWS migration capabilities, but also to engage with us at the strategic level to deliver other IT services. It makes more sense for the migration vendor to understand the long-term IT and organizational strategic needs, especially related to our digital transformation initiative.
Despite the challenges we encountered (and following Eisenhower’s advice), while plans were not as useful, planning proved indispensable. With planning and our hard-working, dedicated, and mission-driven team, we were able to migrate 95% of our IT assets to the cloud and achieve key IT and business objectives.
Saeed Elnaj is an Advisory Board member for HMG Strategy’s 2019 Washington, D.C. CIO Executive Leadership Summit taking place on April 30. To learn more about the summit and to register for the event, click here.