Cloud Migration Roadmap – Transformational Blueprint
The first phase of a Cloud Transformation project determines the principle scope of the exercise based on the business goals, and the second, the Cloud Migration Roadmap, translates these an actionable plan for implementing and achieving them.
Cloud Transformation Roadmap
The primary output of the first phase is the scope definition, which will shape the nature of the required migration work.
For example if it is identified as a purely ‘Lift and Shift’ objective, then this will mainly require a relatively simple process of mapping computing load requirements to their Cloud service equivalents. This is effectively illustrated through this Microsoft case study.
Still within only a Technical scope the migration may also seek to modernize the database layer, like Netlix migrating away from a legacy data-centre Oracle approach to a Cloud Native one of Cassandra and MySQL databases. In this scenario the bulk of the hard work centred around the intensive data migration, not the Cloud hosting component.
The scope can also range through to the wholesale re-engineering of applications and even business models, entirely changing how the organization conducts its business processes, and it could also include a large-scale, enterprise-wide campaign to drive Cloud adoption across the entire organization via one sweeping initiative that encompasses thousands of applications.
It may also be at the other end of the scale, an ultra-simple process where the business needs of a project would be best served by the adoption of a single SaaS application, with no technical migration work needed at all, the effort would entirely be focused on user adoption and training.
Therefore an effective Cloud Migration planning approach must be able to address this entire breadth and depth, and to this end a number of case studies from Microsoft, BT and Credit Suisse are referenced to define the best practices required to articulate a full Cloud Transformation Roadmap.
Enterprise-wide Cloud Solution Design
Microsoft offer themselves as a case study as well as a source of best practices. In this case study they describe their own enterprise-wide initiative to encourage all of their departments to shift away from legacy IT to their own Azure Cloud.
This required moving approximately 2,100 line-of-business applications to the cloud platform, applications that are spread across eight datacenters worldwide, comprising over 40,000 distinct operating system instances.
Microsoft started a central group of Cloud experts, the ‘Stratus’ team, to work collaboratively with departments to build their migration roadmaps, and defined a portfolio management framework where applications are priortized for migration based on a number of key criteria such as the simplicity of the application migration effort and it’s business criticality.
Another Microsoft presentation provides a detailed methodology for this portfolio approach, described in detail as Cloud Solution Design.
Defining a core solution design method of ‘Cloud Pattern Matching’, a process of mapping functionality needs to the appropriate IaaS, PaaS or SaaS option, this would equally address the simple through complex requirements; for example mapping one departments needs to a suitable SaaS option.
Application Architecture Roadmap
As we move into more advanced scenarios, notably the migration of complex enterprise applications, the presentations from Credit Suisse and BT offer the required insights and best practices.
Cloud Solution Design describes a process of indexing business requirements for this mapping process, in a form such as A-1: Customer Portal, and Credit Suisse builds on this same approach expanding it to a full business transformation exercise that links overall Business/IT strategy to the migration roadmap process, creating an Application Architecture Roadmap, described in this OMG presentation: Design Reviews Using the Business Capability Model.
This offers a core reference model for transforming IT solution delivery through Business Architecture, describing a design review process that cascades down through strategic planning and drives roadmap development by breaking these down into the required Business Architecture Capabilities and then mapping these to suitable technical functionality.
Most notably it explicitly describes how Business Architecture bridges the gap between Business and IT, formally establishing governance links between the goals of strategy and the execution of IT, and also how to then implement this linkage across Enterprise Architecture.
- Slide 14 describes the core design process, a process of ‘front to back’ overlaying of the Logical model across the Physical one.
- Slide 13 shows how this produces a Features List, for example T1 – A Trade Store Abstraction Layer.
- App / Data Centre Consolidation – This is part of a simultaneous process of identifying duplicated processes and systems, and eliminating these as part of the migration, enabled by a disposition decision tree.
This includes the required organizational Design Governance models – Credit Suisse explain how they defined one that supports the process of planning legacy migrations through a ‘front to back’ architecture design process that maps out the dynamics between business processes and the apps that ultimately run them.
Architecture Enabled Business Transformation
An ideal companion to the Credit Suisse is How Platforms and APIs Enable Businesses to Participate in the Digital Ecosystems of the Future, a TMF Live 2016 presentation by George Glass, Chief Systems Architect at BT.
Where Credit Suisse defines the Business Architecture and governance framework for driving the development of an Application Architecture Roadmap, BT offers the complimentary technical blueprints.
Naming this as ‘Architecture Enabled Business Transformation‘ on Slide 16, BT explains how a central, very specific Enterprise Architecture governs all new software implementations on to the core BT systems, and how this is primarily achieved through an ‘ACF’ – Architecture Conformance Framework.’
Each new software project is ranked against the ACF, assessing it for conformance with software design principles and system behaviours that yield the desired results, with only those meeting the required maturity level passing through development stage gates.
API Maturity Model
For example projects are compared against assets such as an API Maturity Model, ranking them higher for features like documented UML models, encouraging developers to produce well documented, standardized interfaces. Each is ranked across Levels 1-3 to reflect this increasing maturity, linked to a central EDM – Enterprise Data Model.
Slides 7-9 and 11-13 describe their SDK system – Reusable process blocks and APIs, designed around a principle of utilizing the SOA (Service Oriented Architecture) to enable a COA – “Customer Oriented Architecture”.
Legacy Migration and Consolidation
This assessment stage-gating forms the work pipeline for their Roadmaps, and thus links new product innovation to the central architecture, establishing Architecture Driven Transformation.
It’s an especially compelling presentation when the results achieved from this approach include the consolidation and closure of thousands of applications, where functionality and processes were duplicated across the huge array of legacy IT apps that a large enterprise like BT accumulates over the decades.
Through the API Maturity Model and other central assets, ongoing consolidation of the OSS environment into a ‘Platform’ design has been successfully underway for 10+ years, eliminating over 3,000 redundant applications.
Slide 6 describes “Data driven product launches”, referring to a classification system for new product ideas, and a ‘Product Change Complexity’ assessment framework for tiers 0-6 for new product ideas, categorizing them in terms of software engineering and business transformation complexity.
Implementating Cloud Transformation Roadmaps
SaaS options are available that can be used for planning and managing these transformational migration projects.
- Inputting the indexed requirements as product features (eg. T1 – A Trade Store Abstraction Layer), and prioritizing them. Aha enables you to create scorecards, and use them to prioritize product features.
- Managing dependencies is a very tricky but very important aspect of Cloud migrations. Aha can help by Managing Dependencies Across Releases through Feature Release Synchronization.