According to Gartner, the total spend on Multi-cloud in 2018 was $32.4 Billion, a clear 31% increase from that of previous year. Now, such huge investments need a guaranteed return. In my last article, I talked about what #MultiCloud is and why and why not to go for it. Today, I am going to share my recipe for building an effective Multi-cloud strategy. And I call it ENCHILADA. The concept of tortillas being filled, sauced and wrapped in various forms was clearly defined by the Aztecs. Similarly, by blending the key ingredients of automation, contingency planning, skill management and future-proof strategizing made my recipe transform Multi-cloud strategy from its tortilla state to the whole enchilada.
Below is the list of ingredients for my recipe.
Analysis: Understand the clear necessity for multi-cloud adoption. Is it at all required? Figure out the business purpose, say data storage, application development, etc. for each deployment in your organization’s infrastructure. It’s not necessary to pick and choose one cloud for one purpose, but one needs to have a fair idea on the unfair advantage that the company gets from using multiple clouds for the same purpose.
Automation: In this fast-changing environment, automation is the key. Check the feasibility of a cloud business automation framework for your requirements that can manage multiple cloud deployments end-to-end from cloud billing and provisioning, inventory management, support management, partner management, sales & marketing management to customizable marketplace and business intelligence from a single window.
Contingency: Devise your contingency plans way in advance to overcome common cloud barriers. Work on implementing frictionless governance for reducing risk and ensuring compliance. Also, don’t forget to work on your spending strategies. Replace upfront spends with continuous spends, maximum demand projection with potential use forecasting, capacity management to continuous spend optimization and more.
Emigration: Well, it’s not that we humans alone leave one country and move to other. Even data does so, rather we make data do so. It is certainly not as easy as it sounds, as it depends upon the type of data, size of data, source and destination service provider for migration in cloud and lot more. Multi-cloud is often used for data localization and back-up and henceforth, understanding feasibility of migration is a mandate for any organization before on-boarding.
Design: Spend 80% of your time in making the design correct. Make sure that your architecture addresses streamlined data recovery, critical vendor lock-in avoidance, effective data management, low latency, etc. To maximize the impact of multi-cloud, your architecture must have the capabilities to tackle compliance, migration, security overheads, app sprawl, unique portals and of course, performance.
Harnessing Skills: Identify required skillset for a successful cloud deployment and equip your workforce accordingly. Keep in mind multi-cloud governance (standardized access, structured account, audit trails and others), multi-cloud cost management (cost visibility and allocation, cost optimization and execution, usage and analysis monitoring and forecasting for a minimum viable business), multi-cloud architecture (standards focusing on automated provisioning, operations, recovery, scalability, security and disposability), multi-cloud development (modern development practices with continuous integration and delivery, infrastructure as code, portability, containerization, DevOps and more) and multi-cloud orchestration (leveraging automation and APIs from one or multiple cloud to assemble services) – all of them are equally important and need professional hands to be handled the right way. Hire good cloud architects, DevOps engineers, cost management specialists and project/program managers.
Integration: The difficulties of integration has been evident in every sphere since long. In other sectors, you might dare to survive with this difficulty, but trust me, not in multi-cloud. When selecting cloud service providers, remember considering paths for data integration, process integration and even entire workflow integration, else all your deployments might go for a toss. Finalize your tools for application integration and try them out in real time.
Long Term: Outline a future-proof cloud strategy. Find answers to questions like –
- How many different cloud options will be needed?
- Why to go for #Virtualization or #PrivateCloud?
- Why Multi-cloud over stable on-premise deployments?
With the answers in hand for your current situation, in 5 years and 10 years, plan your own strategy which measures your returns at regular intervals, re-evaluates contracts on Service Level Agreements in terms of quality of service, uptime, performance and more.
Newness: Your organization is unique and so is its problems. Don’t follow others as there is no one-size-fits-all solution available. Be innovative with your approach while keeping the minimum expectations in mind about data visibility, workload portability, software-defined storage, regulation & compliance and others.
With this, I am ready with the secret recipe for building an effective Multi-cloud strategy –
ENCHILADA = Emigration + Newness + Contingency + Harnessing Skills + Integration
+ Long Term + Analysis + Design + Automation
Build your strategy today and stay ahead of others. Try out ENCHILADA and share your thoughts on the same. I will be back with more insights on Multi-cloud soon.