Original Post From Dr. Hossein Eslambolchi
Cloud computing has the potential to drive greater agility, speed and cost savings by providing on-demand access to an elastic pool of shared computing assets—such as applications, servers, storage and networks—that can be rapidly provisioned and released, with minimal service provider interaction, scaled up and down in response to business demands and acquired economically on a pay-per-use basis. In practice, cloud delivery is proving to be all of the above, and increasingly, IT organizations are finding themselves involved in managing hybrid infrastructures that combine public cloud, private cloud and traditional IT services.
Data center evolution is quickly moving beyond service virtualization with expanded requirements for data center consolidation and lowering the cost of complexity. Larger organizations are finding they now need to manage multiple cloud software stacks and configurations across heterogeneous cloud silos. Like mainframes of yore, are clouds becoming the new silos of the 21 century? A unified cloud effort across all IT organizations remains challenging, but it is clear that distributed computing models will again break down the silos and integrate applications across multiple clouds, hybrid clouds.
A cloud strategy starts with making sure your plan will reduce development and deployment cycles times as you migrate applications (virtualize services) to the cloud. Taking a service-centric view and adopting a hybrid cloud abstraction architecture can ensure your developers don’t need to be cloud experts to more rapidly migrate or write apps across hybrid clouds. Your operations staff can modify policies to connect the right cloud for the right workload with the right SLA at the right utility-based price. What is required is a unified set of normalized RESTful APIs for abstracted IaaS that are standardized across an IT development organization and enable freedom of lock-in from the underlying infrastructure. This includes the flexibility to have an object stored in a database one day, then changing an abstraction model to point to a distributed key-value store the next without touching the cloud application. This starts to expose common infrastructure capabilities as Everything-as-a-Service (XaaS) to IT application developers. However, the enterprise still needs to maintain an intelligent linkage of the cloud software stack with business processes and the operational support systems that accompany their critical business systems. This is not a big bang approach to cloud migration, but one that begins with web-based applications and continues with the migration of business–critical distributed applications “one service at a time” to the hybrid cloud.
As cloud becomes more prevalent in IT, further IT governance and standardization of the control plane from automation and orchestration will be required. In fact, development teams expect to greatly reduce development and deployment cycle times by moving to a common abstraction model with a normalized set of RESTful API interfaces. Interfaces to bare metal, to VMware, to OpenStack, to Cloud Foundry, to Amazon, to RightScale to a flexible set of hybrid cloud choices which best suite your organization’s cost, SLA and user experience requirements. As SDN evolves, future cloud resources will be available in “places in the network” as part of end-to-end cloud service delivery as intermediate data processing points for real time M2M data and as SDN control points for application service flow management.
In the new distributed computing model for hybrid clouds, workloads begin to run across one cloud (e.g., private) and scale/burst into another cloud (e.g., private/public/hybrid), with the right infrastructure applied to the right workload (business process). The bursting capabilities span across multiple public providers or multiple private cloud data centers in a cost-brokerage fashion, based on policies that control, or potentially even subscribe to and from a partner cloud broker. The bursting and elastic scale destinations for your big data is enabled with an ease of an overlay system that is the keeper of a library for Amazon RDS, Microsoft SQL Azure, along with a library that includes a combination of other proprietary database or key-value store connectors. Within this hybrid cloud model the enterprise will also require multi-tenancy within a private/public cloud sandbox for zero downtime introductions of new upgrades and releases. For classified data handling, the advanced hybrid cloud architecture must also include “Chinese Wall” capabilities to ensure classified data does not move across country borders, transfer to partners or rest in public/private locations not certified for protecting data privacy and regulatory compliance.
Cloud computing is rapidly evolving and the enterprise cannot afford to wait, but must now begin the journey to achieve immediate cost benefits. However, organizations must adopt a flexible strategy that embraces hybrid cloud architecture and enables a continuous change model to remain nimble across constantly evolving IaaS offerings. Applications should be migrated to the cloud once, with reduced cycle times, with reduced deployment times across an evolving library of abstracted hybrid cloud connectors. Many large enterprises will embrace dual vendor strategies in a cost-brokerage fashion with the freedom to switch cloud providers as the cloud market evolves and shifts.
Companies must shift from their current CAPEX/OPEX investment in private cloud data centers that currently gobble up available budget to run the business. Instead, they must shift budgets into growing their business and begin a migration to a flexible hybrid cloud OPEX-based model for leveraging out-sourced managed private or public cloud IaaS in an “on-demand” or “just-in-time” model. This includes lowering switching costs to enable the freedom of the enterprise to not be locked into a single provider and have the choice to easily switch providers in the hybrid model without having to touch or rewrite their applications to a new set of APIs. The required solution is a new service virtualization layer (distributed service fabric) with a flexible and automated policy overlay that will enable consistent orchestration across heterogeneous and multi-vendor hybrid clouds.
A global policy-based hybrid cloud system enables greater control over cost, performance, user experience and SLA metrics for a variety of different workloads supporting your business architecture (business process). I call this the “Advanced Hybrid Cloud Architecture.” Key cloud capabilities such as elastic scale, fault tolerance (availability), reliability, resilience self-healing, and security must come “out-of-the-box.” These important IT functions should not be reinvented nor recreated by each application developer in silos but provide a consistent set of normalized XaaS APIs that deliver IT governance and consistent control for your organization. However, the hybrid cloud architecture should remain flexible to embrace the customization and extensions required by each IT development group or external ISVs. The new advanced hybrid cloud business and technology architecture must support three IT delivery models for Cloud Migration: Private/Public/Hybrid cloud coexisting with the Traditional Enterprise IT delivery model.
Subscribe to the blog to receive updates about:
AltaFlux understands what you and your organization need to excel, and can deliver rapid innovation to unleash your full workforce potential. Together, we can empower your business by streamlining, transforming, and optimizing your key HCM and talent processes with industry-leading SAP SuccessFactors technology—enabling you to adapt at the speed of change.
AltaFlux Corporation is a global HCM cloud consulting partner based in Troy, Michigan. We empower organizations by streamlining, transforming, and optimizing key human capital management (HCM) processes with industry-leading HCM cloud solutions like SAP SuccessFactors, Benefitfocus, WorkForce Software and Dell Boomi.