The Journey to Hybrid Cloud
Cloud computing has revolutionized IT infrastructure and enterprise applications in businesses from every sector in Qatar. Even government bodies, which have lagged in cloud adoption in many countries, benefited from CRA’s early and proactive publication of cloud policies for departments.
Cloud computing offers a number of core capabilities, namely:
- Dynamic scaling of capacity, deploying more or less resources as the situation demands
- Self-provisioning; customers can create, roll out, and retire services without assistance
- Usage-based pricing, with customers only paying for the resources they use on a moment-to-moment basis rather than across lengthy billing cycles
- Standardized integration, with APIs available to integrate data transfer, management, and security between services
These capabilities offered many customers an irresistible attraction due to the resultant value for money, flexibility of services, agility and rapidity of deploying new capabilities, and scale.
In its early days, public cloud delivered by the outside cloud providers was the only way to achieve these capabilities, either through international hyperscalers Amazon, Google, and Microsoft, or through regional and local internet and telecom service providers.
But public cloud does have its detractors. Concerns about storing sensitive data offshore, and vulnerability to connectivity outages are among the major hurdles. Into that gap stepped private cloud, allowing customers to operate complete cloud infrastructure stacks either in their own datacenter, or within a service provider hosting facility. In partnership with capable technology providers, this delivers nearly all of the promised cloud benefits, with the advantage of closer control of the underlying infrastructure.
Public vs private
Public and private cloud may be similar in the underlying infrastructure, but the use cases are frequently different.
Public cloud is the ideal option for rapidly provisioning cutting-edge services for development and pilot projects, and for the most efficient scaling of resources, especially to very high levels of consumption.
Private cloud is preferred when data sovereignty is essential, keeping the data on the customer’s own infrastructure and not in a third-party facility or offshore. Private cloud may also be a good option for organizations concerned about being locked into a provider’s platform.
The combination of both public and private cloud platforms is effective when migrating enterprise applications into or out of the cloud, providing a staging ground to ensure a smooth migration. And private cloud can be the ideal technology for building cloud-edge applications such as the Internet of Things (IoT) where the application delivery of public cloud is combined with the orchestration and analytics of close-proximity private cloud.
In many cases, it is not that black and white – many of the advantages of public and private cloud can be realized on other platforms, as technology providers mature and expand their portfolios, leading to a combination of public and private cloud: hybrid cloud.
Nearly all customers have been actively deploying cloud services for several years. In fact, the vast majority now consume services from multiple public cloud providers and operate their own infrastructure. We believe that 90% of organizations in the GCC will deploy hybrid cloud and hybrid infrastructure.
The result has been a steady progression to more complex “hybrid infrastructure”, which leverages best-fit solutions according to the needs of separate business units or individual enterprise applications.
While this does result in benefits to individual use-cases, the complexity can quickly become unmanageable. At some stage, IT departments seek to unify management across these platforms, to unify management, provisioning, costing, and security. It is common for CIOs to discover that not only has cloud deployment been uncoordinated, in many cases it is entirely undocumented, resulting in shadow IT which exposes the organization to even greater risk.
It often made sense to deploy individual applications to get them operational as fast as possible. Common instances of this include Customer Relationship Management (CRM), application development, and collaboration applications, but there are many other examples we see every day.
Moving enterprise applications from one environment to another can be a daunting task. It is rarely as simple as just migrating the existing configuration from one location to another. Software vendors and integrators with cloud experience can assist in this process, especially where it involves multiple applications with integration across the enterprise.
Parallel deployment is a common way to reduce risk, deploying an enterprise application into a cloud platform while the existing application remains in production. The new instance can then undergo testing and integration before entering service as a failover resource, then promoted to primary with the previous instance demoted to a fallback role until full deployment is complete. Customers need to focus on reliability throughout this process, but also on efficiency since the longer a migration takes, the greater the risk and cost incurred.
Some applications, particularly those with complex networks of integration and data, lend themselves to a two-stage migration, such as moving workloads from on-premise legacy infrastructure into a private cloud, which allows closer coupling of both infrastructures and minimizes friction. From that private cloud, the workload can then be more easily integrated into a public cloud stack or migrated further.
Regardless of whether the migration is complex or a simple lift-and-shift, that should almost always be followed up by a review of the application to ensure that inefficiencies of the previous infrastructure do not remain replicated in the new location! To fully leverage the advantages of a new platform, especially in the cloud, applications can be rearchitected or refactored to make optimum use of new services and technologies.
Look for platforms and partners which provide tools, best practices and experience in application migration and optimization.
If unmanaged, the key risks of this hybrid approach include:
- A greater expense due to lost economies of scale and duplication of spending across multiple providers
- Security risk, due to added complexity of securing, monitoring and provisioning/deprovisioning access
- Service degradation due to a lack of coordinated management, leading to outages or disruption, or data corruption which may go unnoticed for some time
But this is not an argument against rolling out hybrid services, but an argument for more effective management of it! Organizations should look to retain all the advantages of that mix of platforms and services, while addressing the risks and inefficiencies of poorly managed hybrid infrastructure.
Not only can these risks be addressed, but the organization can directly benefit from a properly managed hybrid environment, utilizing the services which provide the best business impact and value, and then building on that base to achieve genuine hybrid value.
When integrated and managed cohesively, multiple services can be orchestrated to optimum performance, and share automation across applications to expand digital transformation beyond individual silos across the entire organization. By coordinating data management, the stage is set for advanced, cross-application analytics and the future rollout of machine learning and artificial intelligence (AI).
So what does such unified hybrid infrastructure management look like? An effective solution should comprise several layers.
- Technology platforms, across public and on-premise/hosted infrastructure, which enable rather than hinder the journey to optimized hybrid infrastructure
- Integration of solutions to a central management console, and since new services come online all the time, focus on flexibility and extensibility as much as the depth of coverage at first deployment
- Provisioning and deprovisioning, to ensure that as teams change, there is no legacy of forgotten access or, worse, privileged accounts
- Discovery of solutions already deployed, to identify previously unknown cloud applications in use, and move them to a managed state with integrated security and user provisioning
- Telemetry and monitoring, to ensure continuous service delivery to meet KPIs and SLAs, mapped to customer experience (CX) and user experience (CX)
- Security, to coordinate cybersecurity, anomaly detection, behavioral analytics and more
- Data integration, to plan for high-scale growth, reduce complexity, and build foundations for advanced analytics and machine learning
- Managed services, typically covering infrastructure management, application integration, and managed security
Each of these layers involves small steps that can take place in parallel and which offer advantages and efficiencies at every stage without interrupting any applications or services. We suggest starting with a hybrid strategy that provides a clear and ambitious roadmap for taking hybrid infrastructure to the next level, with clear ROI at each milestone and the flexibility to adjust to the changing needs of the business – that’s hybrid digital transformation in action.
GBM has all the required skills and capabilities to support you in your hybrid cloud journey. To find out more on how GBM and IBM can help you, visit IBM Hybrid Cloud Solutions and IBM Cloud Pak for Multicloud Management