Over the last 2-3 years, with the proliferation of digital media and SaaS based technology, it has been rather difficult for IT teams to keep applications (e.g. CRM, collaboration tools, mailing applications, etc.) away from the enterprise firewall. Today, the concept of leveraging applications and infrastructure from multiple cloud vendors and CSPs (across SaaS, PaaS and IaaS offerings) has become more mainstream. The ‘multi-cloud’ construct has now evolved to encompass even hosted environments, private cloud and even on-premise data centers. According to IDC, 85% of enterprise IT will be classified as multi-cloud by 2018, with the top 5 IaaS / PaaS vendors will have at least 75% market share.
In the context of enterprise technology, having a multi-cloud strategy is becoming essential, due to the following reasons:
- The rapid pace at which technology is evolving, making it very difficult for traditional enterprise IT models to keep up. For example, we are seeing rapid changes in enterprise applications such as ERP, HRM, Core Banking, Business Intelligence, etc. Such applications traditionally changed very little, with typical upgrade cycles occurring once a year.
- The convergence of consumer technology with enterprise applications, making business services more accessible and less controlled. For example, the ability to access reports and analytics on mobile phones and tablets, over the internet, was something unheard of even five years ago.
- The exponential growth of workloads, in terms of volume as well as types, makes it very difficult for CIOs to make accurate projections of storage, network and compute needs. Organizations are better placed when they create a flexible technology roadmap and keep their choices open, in terms of infrastructure and applications.
However, most organizations today, even those with best-in-class IT sourcing, deployment and management, are still trying to come to terms with the variable nature of the multi-cloud. In many ways, this new model contrasts with the earlier approach of structured and approval based IT provisioning. Here are five steps that CIOs need to follow to ensure that their organization has an effective transition to a multi-cloud strategy.
A. Simplify Access to Business Users
How the multi-cloud strategy differs greatly from typical enterprise IT models is that technology selection and adoption is often initiated by business-side / departmental users. Let us take the example of a CRM application for a consumer goods company. Typically, CRM deployments would be centrally managed by the IT team. The choice of product would depend on inputs taken from various product teams, cost constraints, customization needs, infrastructure limitations, etc. This process occurs once, during the requirement gathering and customization phases. As markets and sales strategies evolve quickly, certain departments may have new requirements, which are much more difficult to incorporate later on.
In today’s world, since the pace of market change is very high, departments need to have the ability to access and provision applications and infrastructure quickly, without having to go through highly regimented approval, customization and provisioning processes. This is where IT, rather than play the role of approver, needs to simplify access and provisioning. They need to provide business users with a simple and straightforward layer of abstraction that maps all internal and external IT resources, so that departments can get moving quickly. At the same time, IT needs to continue creating and managing security and governance protocols, to ensure that all IT initiatives align to larger organizational goals.
B. Map Workloads to Resources
The concept of workloads becomes very important when you need to make the best use of cloud infrastructure, platforms and applications. All major cloud platforms and Cloud Service Providers (CSPs) have a wide spectrum of services catering to different types of workloads.
Implementing a multi-cloud strategy requires a high degree of clarity around different workloads within the organization, and their individual needs around scalability, performance, security, network load, etc. IT teams need to create a detailed map of current workloads as well as future scalability and performance needs, and then map these to different types of cloud vendors and utilities.
It also allows IT teams to classify different types of workloads, e.g. high volume transactional workloads, mass storage, business critical applications, peripheral applications, etc. Such classification allows IT to map workloads to specific cloud vendors and services that best suit these requirements.
C. Focus on Use Cases and Business Services
With organizations and teams transitioning to a multi-cloud environment, the role of the IT team will also change dramatically. IT personnel will have to work closely with various departments / teams to identify use cases and define business services.
Some of the core areas where IT will need to focus on are:
- Creating and managing a flexible services architecture that incorporates a wide spectrum of cloud services – including PaaS, SaaS, IaaS, datacenters, private clouds, hybrid clouds, etc.
- Understanding the relationships between business processes and workloads, in order to create business-driven technology roadmaps
- Discovering opportunities to optimize performance, availability and cost, by availing multi-vendor and hybrid deployment options
- Implementing effective data governance, risk mitigation and security protocols across a wide portfolio of cloud-based services
Focusing on use cases and business services allows organizations to start small, one service at a time, without having to make major investments in time and resources. The low risk factor motivates teams to create Proofs of Concept and trial instances, paving the way for large scale adoption of cloud based services.
D. Create an Initial Multi-Cloud Governance Mechanism
While it may not be very detailed and optimal initially, it is essential to put in a governance mechanism right at the outset, when you are starting on your multi-cloud strategy. Some of the important aspects of this governance mechanism should be:
- Data governance, privacy and security norms
- User types and access rights
- Processes / workflows for cloud resource and managed services provisioning
- Cloud service requisition processes
- Workload types and Service Level definitions
- Provisioning norms for various cloud platforms (Netmagic Public Cloud, Amazon Web Services, Microsoft Azure Public Cloud)
- Application and data migration processes and controls
E. Use a Cloud Management Platform
Eventually, a multi-cloud environment is likely to end up extremely varied and complex. Decision makers on both the business and IT side will face many challenges in standardizing, managing and governing all the cloud resources being used across the organization. Also, organizations will continue to depend on a wide variety of legacy, in-house applications, data and technology infrastructure. The multi-cloud environment will have a large number of integration points with these in-house applications and data storage.
CIOs need to implement a comprehensive Cloud Management Platform to ensure end-to-end visibility across the spectrum of cloud resources, and enable better provisioning, monitoring and management of services across the organization.
Taking a multi-cloud approach gives organizations unmatched flexibility in choosing IT resources for evolving business needs. While moving towards a multi-cloud strategy, organizations will face a number of unique challenges around legacy applications, integration issues, business process complexity, role definitions and setting clear objectives / KPIs. Of course, every organization needs to develop its own multi-cloud roadmap, depending on its unique business challenges. However, keeping the above steps in mind will help them define a structured approach to multi-cloud adoption and management, early on in their cloud journey.