Sailing the cloud wave

Cloud computing is fast becoming the go-to model. Many enterprises have already adopted software as a service (SaaS) for their email, HR and CRM functions. Some have also started experimenting infrastructure as a service (IaaS) and shifting storage and computing to the cloud.

The cloud brings inherent advantages like on-demand provisioning, resources pooling and elasticity. While there is fear of missing out (FOMO) amongst business and IT leaders, there are many lingering questions also in their mind.

- What will happen to enterprise owned data centers?

- Is shifting all workloads to the cloud is advisable?

- Are early doubts on security and latency of public cloud unfounded?

- What is the real cost of the cloud and long-term cost effectiveness?

- Who owns the data and what happens when provider goes out of business?

- With 5G and Edge, is cloud computing already a passé?

Answers and its significance vary from enterprise to enterprise. Moving-to-the-cloud strategy must be driven by an enterprise’s own current state, business goals and priorities.

Following the 5 Ds based on simple life lessons can be guiding lighthouse in this journey.

1. Discover — Know your landscape

“There is no substitute for accurate knowledge. Know yourself, know your business, know your men” — Lee Iacocca.

No second guessing on this. Have your comprehensive workload inventory up-to-date. Categories it on the scale of core to non-core based on data security & control requirement, latency threshold, application state, elasticity needed and modernization plan. This will be your golden source.

2. Define workload deployment model. Rent-it or own-it?

“Use your worry, but know when to leave it behind” — Twyla Tharp

Define best suited deployment policy considering type of control your enterprise wants over data and operations. It could be combination of own data center, public, private and hybrid cloud. Private cloud also has colocation and on-premise variants. Hybrid cloud provides security and control of a private cloud with scalability and resources of a public cloud.

Large enterprises do carry legacy application and data not suitable for the cloud. Same time they also have redundant application inventory and hardly used data archival.

For startups and small businesses, data security concern may not be that critical. Public cloud can provide them jumpstart. Advantage of opex over capex is strikingly beneficial to them.

3. Decide cloud right strategy. Which cloud provider for which service?

“The choice you make in life will make your life. Choose wisely” — Michael Josephson

With cloud providers offering unique set of services, an enterprise would like to have the freedom to run their workload on any cloud. Key driver behind multi cloud is to avoid sole vendor lock-in or to take advantage of best-breed cost optimized solution or to meet specific data sovereignty requirement. Multi cloud could be regulatory requirement soon.

Nonetheless, it increases the difficulty of integrating cloud services with existing infrastructure and adds extra layer of management complexity. You need single pane visibility to monitor and control all.

Finalize enterprise level cloud providers and create framework to manage their contracts judiciously.

4. Do it on ground. Devil is in the details.

“The real destination lies in step taking” — Ernest Agyemang Yeboah

Final decision should always be left at application level. Every practical scenario should be evaluated during execution. Listing some for illustration.

- For deployment model zero-in, clarity on data repatriate contract with provider is important.

- Analyze cost effectiveness of lift and shift vs cloud enablement first option

- For stable legacy application, why not continue with current data center unless it is a bottleneck for growth plan.

- If the workload is predictable with no major fluctuation, should you prefer private cloud or continue with data center?

- Where modernization on the card, the application should be built using containers and microservices which will be default cloud ready and vendor agnostic.

- Why custom build when better SaaS solution available?

5. Know the downside and be aware of upcoming developments.

“Change is the law of life. And those who look only to the past or present are certain to miss the future” — John F Kennedy

Technology — Cloud computing has centralized way of operation. With more workloads moving to the cloud, more network roundtrips needed to access, process and analyze the data. This roundtrip creates three main issues: latency, security and bandwidth.

Edge computing promises to decentralize resources and bring data and computing power near the source and where it is needed most. Thus, reducing latency and bandwidth issue.

Cost — While cloud’s pay per model is convenient, it is easy to lose visibility and control over costs along the way. Cloud FinOps is big watch out item to optimize cloud spend.

Industry vertical cloud — End to end industry vertical offerings bundled with templates, data sources and connectors as PaaS or SaaS are next big thing.

Bottom-line:

The cloud is an unstoppable force. As IT and business embrace new ways of doing things, now is the time to rethink your strategy, structure, and governance to maximize the long-term benefit and also to prepare mitigate any possible risks of cloud inefficiency.

Move fast but not too fast.