Analytics cloud services     

Analytics cloud services - overview

Overall Vision and Scope:

The IT industry is currently witnessing two important trends that that are fundamentally changing the way enterprises do business in the smarter planet era – the emergence of cloud computing technology and the rapid growth of digital information from trillions of interconnected internet of things. With a tightening IT budget, the “Cloud” is quickly becoming the preferred computing method across organizations because of its inherent advantages like rapid time-to-value, pay-as-you-go model and reduced capital/operational expense requirements. On the other hand, the emergence of “Big Data” is driving IT requirements for timely and comprehensive analysis of a deluge of digital information from multiple sources that are rapidly growing in volume, velocity, variety and veracity.

Enabling efficient Big Data capabilities on the “Cloud” requires smart and automated end-to-end selection, deployment, monitoring, and run-time management of a large pool of IT resources, that can potentially run into hundreds and thousands of machines. Ensuring such applications meet users’ Quality of Service (QoS) targets is not addressed by traditional Cloud computing providers where the primary focus is migrating generic web-based and traditional enterprise application to the Cloud. Moreover, the traditional Hadoop IaaS cloud providers are low-level; they leave the decision of IT resource scaling to the user, thereby making development of Big Data applications difficult and resource provisioning unintelligent and inefficient.

To address the above issues, IBM Research is driving the development of an industry-leading agile, simple and flexible Analytics Cloud, as illustrated in Figure 1. It will continuously ingest “Big Data” flowing in from various sources like telecommunication service providers, video surveillance, multiple sensors, social media, etc. and will be the foundation of smarter solutions services.

Figure 1: Analytics Cloud as the foundation for Smarter Government and Industry Solutions Services

Proposed Architecture

Figure 2 illustrates the proposed high level architecture of Analytics Cloud. Several key IBM differentiating capabilities will be part of this architecture; they include a world-class dynamic large-scale IT infrastructure orchestrator and multiple middleware assets for Big Data ingestion and analytics for prediction, optimization and visualization. IBM Research is partnering with the business units and building several key innovations to ensure rapid onboarding and timely delivery of various government/industry solutions and services.

Figure 2: Architecture of Smarter Government and Industry Solutions Services delivered on Analytics Cloud