Recent Contributions

Keywords:Cloud computing, internal cloud, virtual machine, cloud framework
Authors: Alexander Schmidt, InRule Technology, Inc. and Dmitry Tkach, Schneider Electric.
Abstract:
This paper discusses the approaches for a Virtual Machine optimized for executing in Cloud environment. The optimization concerns with I/O operations, memory management, and job scheduling. We address the problem of development of software using the provided Cloud environment. Our first contribution is the re-modeling of existing frameworks for traditional distributed storage systems. The second contribution is the introduction of job scheduling framework as a standard component in the system for developing event-based solutions; since the resulting framework is generic, we show how to use it in managing the data through sample problems. Consequently, we contribute to the design of a library of proof-based developed algorithms.

Keywords: analytics, data mining, developing world, macroeconomics, digital divide
Authors: Ajay Ohri, India.
Abstract:
The cloud computing paradigm offers unparalleled access to computing resources both in terms of storage as well as processing power for developing countries.The use of predictive analytics and data mining has been hitherto restricted to an elite set of universities and organizations willing to invest tens of thousands in annual license
fees to software companies like SAS ,SPSS,Oracle and SAP and even more in terms of network and server hardware costs to companies like HP,Dell and IBM.Every two or three years, the hardware needed to be upgraded , thus putting total cost of ownership of predictive analytics, data driven decision making and resource planning well out of reach of a major part of the planet's population.Copyright infringements and intellectual property violations further helped create a divide between advanced computing and those who needed it the most. Now thanks to open source softwares, softwares as a service and cloud hosted processing , even a relatively non funded Indian or Asian or African university ,government office as well as small and medium enterprise can avail the advanced cost savings due to predictive analytics. This in turn will lead to a new era of resource optimized decision making, one which benefits all companies that offer the flexibility of cloud hosted applications to hitherto closed markets.

Keywords:platform, operating environment, infrastructure
Authors: Rob Gingell, Cassatt Corporation.
Abstract:
Cassatt Active Response (CAR) is a family of products that permit the pooling and dynamic configuration of processing, communication, and storage resources. Configurations are established to support the execution of multi-component applications, within and across data center and organizational boundaries and are automatically adjusted to meet service levels, respond to failures and changes in capacity, and to add and remove new services as needs change.

The discussion will both describe the system and its evolved requirements as well as experiences in applying it to a variety of environments. The discussion will focus on the "why" it does what it does in addition to the "how", both to motivate the approaches taken as well as convey attributes essential to any approach to cloud computing. The discussion should be useful to those contemplating trying to create or modify their datacenters and resources to become more fluid in response to the class of dynamic demands expected to be typical in cloud-like usages.

Significant aspects of CAR's design are the high degree of both resource and application heterogeniety and the systems network-centric, loosely-coupled management of the environments it constructs and maintains. The system supports most major hardware and system software families, a large variety of networking hardware, and popular storage offerings and is suitable for the variety of resources typical in many data centers. More importantly, the system supports a broadly heterogeneous notion of logical resources extending from physical entities through containers such as virtualized hardware, and abstract virtual machines such as operating system processes and interpreters such as the JVM and CLR, and virtualization technologies such as VMware, Xen, Parallels, and Solaris Containers.

The system provides a common and consistent model for the composition of stacks of software within "machines" and containers, and also among distributed such stacks and across data center and organization boundaries. Its augmented model of containers creates uniform points of reference and measures of resource consumption, capacity, and utilization efficiency which guide its execution and provide the basis for access control, billing, and capacity planning and management.

CAR can employ the resources of other pools as in external "cloud"-like offerings from Amazon, and can offer its capacity in return and is a tool that can be used to construct so-called "internal clouds" and to manage application deployments across resources in both internal and external cloud environments.

CAR does not directly provide an application environment, instead providing the infrastructure on which the wide and increasingly diverse collection of application frameworks can be supported. It provides the resource allocation, arbitration, and management functionality which is usually missing from such frameworks. It is a basis for supporting the emerging class of developers charged more with "deployment" than "coding", as applications continue their trend towards a greater population of common components distributed across a variety of services and environments.

Keywords:cloud computing, enterprise software, software licensing, private clouds, cloud-based software licensing
Authors: Stuart Charlton, Elastra Corporation.
Abstract:
Still a relatively new technology, cloud computing has evolved rapidly in the past two years. The opportunity to enable a new form of enterprise software licensing is now a reality and this approach boasts promising rewards. The benefits of cloud-based software licensing are relevant to all Enterprise IT applications. Such systems continue to grow its complexity in both businesses and in the software they consume. Enterprise software product licensing has been limited to desktop deployments, creating limited flexibility for software vendors. Enterprise cloud computing removes the fundamental limitations to issuing customized licenses for users and harmoniously linking developer-centric cloud platforms with ecosystem-driving cloud infrastructure.

The roadmap towards this technology is being executed today: an ecosystem for the cloud is rapidly forming, supported by digital identity technology, a licensing markup language, and cloud server reference architecture to integrate license enforcement with the overall security in the cloud. Elastra’s vision for the future of cloud computing is clear. Stuart Charlton will outline a blue print for cloud-based software licensing and what that means for public and private clouds in the enterprise.