Application Refactoring & Migration

ENGAGEMENT

One of the largest telecom providers in North America needed to centralize private cloud capacity as part of an effort to consolidate regional data centers into their national data center footprint to simplify operations and reduce operating costs. Consolidating data centers required extensive application migration planning and execution. DayBlink Consulting was engaged to design and orchestrate the migration of 400+ applications including core infrastructure service applications, customer-facing applications, and numerous private and public cloud hosting solutions.

PROBLEM

The Cloud organization needed to evacuate a regional data center in less than a year while ensuring business continuity as many of the hosted applications directly supported customer-facing applications. There were several challenges that significantly increased the complexity of the migration: (1) there was no reliable source of truth for an application inventory, application owners, application-specific requirements, cloud platform requirements, or application interdependencies, (2) the applications and services lacked clear destinations due to infrastructure availability, security limitations, and budget, (3) the team had no solutions architect with institutional and cloud knowledge that was able to designate clear migration plans, (4) planning and coordination spanned over 80 application teams involved in the migration responsible for migrations, and (5) all migrations were to be completed in advance of a looming data center lease expiring and strict decommission date.

SOLUTION

We were engaged to migrate all applications in the regional data center so the organization could shut down the facility. We needed to understand what applications were in the data center, develop tailored application migration plans, migrate and validate application moves, and remove the application footprint from the source infrastructure. 

This migration took shape through the following phases: (1) conducted a discovery effort to develop an inventory of applications, application owners, application-specific requirements (e.g. geographic requirements, footprint, etc.), and interdependencies between applications, (2) evaluated application requirements and determined target platforms and locations, (3) prepared tailored migration approach fit for each application and service needs (e.g., re-platforming, re-factoring, lift-and-shift, etc.), (4) analyzed historical and forecasted application usage and mapped to confirmed capacity availability at various target data centers and platform locations, (5) developed a phased migration approach across all application to accommodate application interdependencies and ensure there were no connectivity impacts, (6) facilitated the migration and onboarding of applications and services to deployed private and public cloud infrastructure, and (7) removed application footprint from the source infrastructure and validated operational continuity. Throughout the migration, we launched and managed an awareness campaign to ensure clear and consistent communication (e.g. office hours, dedicated messaging channels and spaces, 1-on-1s, intranet update, T-minus email broadcasts,  team-specific check-ins, and more).

A key pillar to ensuring a successful migration involved determining application-specific migration strategies that aligned to the organization’s constraints and long-term business objectives. After gathering critical details including application requirements, available private cloud capacity, each application’s public cloud footprint, and the organization’s technology roadmap, we aligned on a migration approach (e.g., re-platform, re-factor, lift-and-shift, etc.) fit for each application.

For the applications migrating to another on-prem environment, we identified the target on-prem locations to shift the workloads to. For many applications, there wasn’t enough available private cloud capacity so a new on-prem environment had to be built out and eventually the workloads were shifted to the new location. In addition applications migrating to on-prem locations, many others pursued an on-prem to cloud migration. Most of these applications were either re-factored, re-platformed, or re-hosted (i.e. lift-and-shift) depending of factors like estimated level of effort to re-factor vs re-platform, available budget, and technology roadmaps (i.e., future plans to replace applications in the current ecosystem).

Eventually, we sequenced all on-prem and cloud migration plans with a clear path to exit the source environment with migration interdependencies documented and tracked. Once we shifted all application workloads to the target environments, we systematically decommissioned the source environment while monitoring for any operational impacts.

RESULT

We collaborated with 80+ application teams to successfully migrate and exit 400+ virtualized applications, 12+ bare-metal applications, and 3+ partner organizations from the regional data center without impacting business continuity or uptime. 

Application teams migrated and onboarded to new cloud hosting solutions including AWS, VMware, and OpenStack.

A new “source of truth” application database which included application profiles to capture application descriptions, owner information, dependencies, requirements, and more. 

A comprehensive strategy & approach playbook documented to support and streamline future data center migrations as the company continued to consolidate its cloud capacity.