Simplifying cloud adoption is one of the surest ways for businesses to help IT become an innovation driver. It's also common sense given the technology's role in daily office life and large-scale digital transformations. Nonetheless, doing so requires architects and business teams to stage an impressive act of bottom-up and inside-out change initiatives.
Gartner Research estimates that 45% of IT spending in 2024 will go towards cloud-based solutions. These investments support technological and organizational changes for enabling and continuously improving digital business models — the likes of which are contingent on cloud infrastructures.
Innumerable market factors are behind this trend, but since last year, the COVID-19 pandemic is giving concrete proof that cloud-hosted services can deliver resilient, scalable, and re-usable IT landscapes. In fact, 2020 was the first year that enterprises put more money into cloud infrastructure services than on-premise data centers and hardware, a victory for cloud evangelists stemming from years of stakeholder buy-in and value proof points.
Unsurprisingly, business leaders want closer ties to enterprise architects (EAs) and those throughout the IT value chain as their digital business models come to fruition. Their achievements and failures, as well as their methodologies and tools of choice, are evaluated more than ever before in the context of business outcomes. Cloud migration roadmaps are expected to be crystal clear yet also adaptable, complex IT models must be distilled into digestible formats for new varieties of users, and IT portfolios must be innovated upon and their costs rationalized at the same time.
What’s lesser-known to business audiences are the requirements for implementing and governing decentralized IT landscapes. To put it in perspective, on-premise applications en route to the cloud come with heavy technical baggage. The slightest change to their functional makeup can disrupt interconnected business processes and formal procedures related to compliance. Trade-offs must inevitably be cast when migrating these services, and if a transition is not tied to a concrete business objective, decisions will be delayed — if made at all.
As cloud adoption occurs, governance over multi-cloud, single-cloud, or hybrid IT environments composed of microservices must account for speed and quality. This means implementing and monitoring custom policies, guidelines, and guardrails for cloud usage to ensure services are procured and consumed rationally. Shadow IT is a threat even in the cloud, and unless companies use automated methods to see where and how decentralized services run, there is a high risk of infrastructure sprawl. Indeed, such actionable transparency is the only way infrastructure, operations, and governance teams can drive collaboration and knowledge sharing during cloud migrations efficiently.
12 steps for cloud migration and governance
The cloud and breaking down monoliths into microservices represent two of the most important evolutions of technology. But as emphasized at LeanIX’s recent EA Connect Day APAC by Kati Gholam, enterprise architect at TUI, architects must be prepared to transform their organization while modernizing IT. A cloud and microservices journey is no different. However, to create an organization and culture that can support these technologies, EA teams and SaaS Centers of Excellence (SCoE) must unlock end-to-end transparency throughout all stages of the cloud migration and governance process.
Assess cloud-readiness
Define target capabilities, decide which applications will move to the cloud, and which applications should remain on-premise. At this stage, it is necessary to gather information on applications, IT components, and interfaces related to business criticality, security, and monitoring needs.
1. Build inventory and collect data in survey- Run a survey across your organization to gather application data regarding the planned migration strategy
- Align TIME model with 6R strategy and decide the migration path for your applications
Plan and execute migration
Identify all viable business alternatives associated to any proposed cloud technology when transferring IT assets to the cloud. Evaluate cloud transformations based on explicitly-detailed degrees of impact, estimate migration timeframes, and calculate the budget and ROI of changes.
3. Model migration strategy (6R) as impact- Employ the 6R strategy:
- Refactor (rewrite applications to become cloud native)
- Rehost (migrate without making further changes)
- Repurchase (switch to a new service and licensing model)
- Retain (keep as-is until a later date)
- Replatform (alter select elements for the cloud)
- Retire (remove from portfolio altogether)
- Analyze dependencies to other projects and epics
- Select transformation plan
- Jump to timestamp
- Visualize changes in heatmap
Establish cloud governance
Manage cloud resources with an up-to-date repository on cloud components integrated directly to major cloud providers. Whether for cloud portfolio management, cloud sourcing, cloud spend showback, and cloud governance and compliance, automation is the only way to gain transparency over decentralized landscapes.
6. Discover multi-cloud services across AWS, GCP, and Azure- Provide business context to cloud services
- Mitigate business-relevant violations
- Allocate services and showback cloud spend
- Enable account owners to adhere to standardized tagging
- Set up the basis for comprehensive governance
- Assess balance of cloud spend vs. delivered business value
- Detect unallocated resources and optimize spend
- Determine violation types based on security, availability, best practices, performance, obsolescence, and cloud spend
Catalog microservices
The visibility of a microservice catalog saves DevOps teams time when navigating microservices and automating documentation. The success (and ongoing risks) of deployments are available via clear metrics and resources can be allocated to the actual development of products.
10. Detect microservices from CI/CD and Kubernetes- Sync metadata from CI/CD pipelines and orchestration platforms into microservice catalogs to more efficiently:
- Navigate and operate along the microservices landscape
- Mitigate business-relevant violations
- Onboard new developers quicker and enable support function
- Proactively manage open-source licenses
- Monitor deployment KPIs to increase development efficiency
- Track metrics such as deployment frequency, mean time to resolution (MTTR), and failure rates
- Encourage self-education
- Eliminate impediments to knowledge exchange
