Service Oriented Infrastructure
Service Oriented Infrastructure is a systematic means for describing Information Technology infrastructures in the terms of a service. The underlying principles of Service Oriented Infrastructure revert back to LDAP and Mainframe technologies, among others. Where Service Oriented Infrastructure manages to distinguish itself is the way in which it provides a mindset or framework for making Business benefits measurable.
Service Oriented Infrastructure tends to provide the foundation for Information Technology services. Intel initially developed the concept that had three different domains for Service Orientation; they were the Enterprise, the Infrastructure, and the application architecture.
For now, we will focus on the infrastructure part of that equation.
Vital aspects of Service Oriented Infrastructure include Virtualization and Industrialization. These components provide Information Technology Infrastructure services through a pool of resources, rather than through discrete instances.
While Service Oriented Architecture tends to be widely adopted throughout the Information Technology industry, the adoption of Service Oriented Infrastructure has been a lot slower. This is beginning to change, however, with the wide availability of such Service Oriented Infrastructure solutions as database grids, virtualized storage, application server grids, and virtualized servers.
The following definition of Service Oriented Infrastructure has emerged from a joint effort between Capgemini, HP, and Cisco: a virtualized Information Technology infrastructure that consists of components that are typically managed in an industrialized fashion and which are comprised of Service Oriented Architecture Application Support and manage to expose a catalog of services, rather than discrete instances.
The phrase Service Oriented Infrastructure has a much broader usage, however, which includes many configurable infrastructure resources, including storage, computing, and networking software and hardware that effectively support how applications are run. Service Oriented Infrastructure is consistent with all Service Oriented Architecture applications. Service Oriented Infrastructure works to facilitate the reuse and allocation of infrastructure resources.
A Service Oriented Infrastructure exposes a set of fundamental services that might include security or mobility, services that are part of a network environment that are able to deliver resource sharing, collaboration, communications, and application integration that is scalable, maintainable, ubiquitous, reliable, and also cost effective.
In April of 2007, the Open Group had a project on Service Oriented Infrastructure within its Service Oriented Architecture Working Group. The goals of the Service Oriented Infrastructure project were to develop a more complete understanding of Service Oriented Infrastructure.
The Orchestration of Virtualized Components
Service Orientation is unique in that it provides a lot of tremendous advantages for Information Technology Infrastructure services. The primary benefits include an increase in the utilization of individual resources as well as an increase in service levels, since applications are not dependent on the availability of individual resources and can thus utilize any single resource that is available within the entire pool.
The Information Technology Infrastructure Technologies that are available in today’s market provide numerous options for the deliverance of high quality end to end Service Oriented services. Every service within such a domain may be virtualized through Schedulers. The required quantity of resources for the constitution of a service may be managed through a highly automated provisioning process that ensures high quality standards while simultaneously enforcing the consistent behavior of the Infrastructure Services. This can be applied to servers, databases, storage, directory services, and networks – for every aspect of the Information Technology infrastructure.
A Service Oriented Infrastructure Road Map
A three step model has been deployed by Capgemini in order to reap the full benefits of Service Oriented Infrastructure in a Business setting. They are as follows:
- A three day Service Oriented Infrastructure maturity scan to assess the current state as opposed to the best in class Service Oriented Infrastructure implementation.
- A three week mobilization and strategy plan in order to designate the road map for the development of a Service Oriented Infrastructure based ICT Infrastructure, as well as to mobilize the company around the plan
- A transition and implementation engagement to construct and design a solution utilizing Cap Gemini’s Infrastructure Design Framework.
Service Oriented Infrastructure Frameworks
A Service Oriented Infrastructure can be thought of as a flexible and modular Information Technology fabric that based on standard building blocks that come highly configurable to meet requirements that continue to evolve at a dizzying paced. Regarding service oriented solutions in terms of multi layered structures, the Service Oriented Infrastructure is focused on the virtualization and orchestration of storage, network and computational resources.
The Service Oriented Infrastructure guarantees that resources will be made available in the location and quantity required by the Service Oriented Architecture tier that is situated above it. Within the Service Oriented Infrastructure abstraction, a device’s physical details may be hidden by services on the platform. The device is then capable of being managed via abstract service interfaces that are specific to SLA.
Service Oriented Infrastructure is optimizable for handling a high amount of XML traffic that is typically associated with web service applications. Service Oriented Infrastructure also utilizes XML as a lingua franca for enabling the interoperability of security services and management built in to each of the building block components. The Service Oriented Infrastructure provides a means for managing computing resources in lockstep with application requirements that are both situated at initial deployment.
On a more specific level, the Service Oriented Infrastructure level of a service oriented enterprise is meant to handle the following tasks:
– Orchestration
Orchestration entails the management of hardware as a set of distributed resources that are also fungible to a certain extent. This represents a shift from a static paradigm to a dynamic provisioning that is based on real time activities and work loads.
– Asset Discovery & Management
This entails the maintenance of an automatic inventory of all devices that are connected. It is always updated in a timely fashion.
– Provisioning
This entails the enabling of bare metal provisioning, the coordination of the configuring among storage, network and server in a synchronous fashion, making sure software gets loaded on to the correct physical machines, taking platforms in and out of service for the testing, repair, capacity expansion and / or maintenance; while simultaneously providing for the remote booting of a system from another system, as well as the management of licensing typically associated with the deployment of software.
– Virtualization
Virtualization makes it possible to run several applications while sharing one physical device as a means of increasing user rates, or allowing for the allocation of multiple machines and storage devices to one application as a means for increasing the performance rate.
– Load Balancing
This involves the dynamic reassigning of physical devices to applications as a means of guaranteeing that pre-specified service levels will be adhered to and the utilization of resources shall be optimized as workloads evolve.
– Capacity Planning
Capacity planning deals with the tracking and measuring of the consumption of virtual resources. This enables planning for the reservation of resources for specific workloads, and specifies when new equipment needs to be brought online.
– Utilization Metering
This entails tracking the utilization of specific resources as designated by SLA and management policy. Metering services could be utilized for charge back and billing by software on a higher level.
– Monitoring and Diagnosis of Problems
This process verifies that virtual platforms remain operational. It also detects network attacks as well as error conditionality, and responds to such situations through running diagnostics and deprovisioning platforms while reprovisioning the affected services. Alternately, it may also isolate network segments as a means of preventing the spread of malware.
– Security Enforcement
This process enforces automatic device as well as software load authentication. It works to trace access, identity, and trust mechanisms within and across corporate boundaries as a means of providing secure services across firewall systems.
– Logical Isolation and the Enforcement of Privacy
This component ensures that a fault in a virtual platform does not wind up propagating to another platform in the same device, as well as that no data leaks are able to take place across virtual platforms that might belong to separate accounts.
– Information Technology Operations Processes
This sets up generic micro Information Technology operations as building blocks as a means of standardizing Information Technology processes while enabling interoperability to take place across heterogeneous system management products.
Service Oriented Infrastructure can provide an abstraction service layer to Service Oriented Architecture applications. It should be noted that the back end Information Technology infrastructure management also must be service oriented as a means of managing and supporting both infrastructure and applications.
Service Oriented Management is an integral segment of the service oriented computing infrastructure that gives management services as well as interfaces to engineers working in the Information Technology field. Platform management can be delivered through what is known as a Standard Platform Interface, or SPI, which is based on such open standards as WS Management.
