An SOI is a modular, flexible IT fabric based on standard building blocks that are highly configurable to meet rapidly evolving requirements. Looking at service-oriented solutions as a multi-layered structure, the SOI layer focuses on the orchestration and virtualization of compute, network, and storage resources. The SOI ensures resources are made available in the amount and location required by the SOA layer above it. Within the SOI abstraction, the physical details of a device are hidden by services on the platform. The device can then be managed through abstract, SLA-specific service interfaces.

SOI is optimized to handle the high-volume XML traffic associated with Web-services applications. It also uses XML as the lingua franca making possible the interoperability of management and security services built into each of the component building blocks. The SOI provides a way to manage computing resources in lockstep with application requirements both at initial deployment, and as the workload or requirements change, effectively enabling an integrated design, deployment, and management life cycle. The standardized and loosely coupled nature of SOI also reduces complexity and increases the potential for automation by enabling devices to diagnose and repair themselves, with minimal involvement from higher levels.

More specifically, the SOI layer of a service-oriented enterprise manages the following tasks:

• Orchestration: Managing hardware as a set of distributed and to some extent, fungible resources, shifting from a static, "one-application-per-box" paradigm to dynamic provisioning based on real-time workloads and activities. This provides the ability to realign compute, network, and storage resources as needed.
• Asset discovery and management: Maintaining an automatic inventory of all connected devices, always accurate and updated on a timely basis.
• Provisioning: Enabling "bare metal" provisioning, coordinating the configuration between server, network, and storage in a synchronous manner, making sure software gets loaded on the right physical machines, taking platforms in and out of service as required for testing, maintenance, repair or capacity expansion; remote booting a system from another system, and managing the licenses associated with software deployment.
• Virtualization: Making it possible to run multiple applications sharing one physical machine or storage device to increase utilization rates, or allocate multiple machines and storage devices to one application to increase performance. In other words, one-to-one dependencies between applications and platforms are removed. This capability provides unprecedented flexibility in meeting SLAs.
• Load balancing: Dynamically reassigning physical devices to applications to ensure adherence to specified service (performance) levels and optimal utilization of all resources as workloads change.
• Capacity planning: Measuring and tracking the consumption of virtual resources to be able to plan when to reserve resources for certain workloads or when new equipment needs to be brought on line.
• Utilization metering: Tracking the use of particular resources as designated by management policy and SLA. The metering service could be used for charge back and billing by higher-level software
• Monitoring and problem diagnosis: Verifying that virtual platforms are operational, detecting error conditions and network attacks, and responding by running diagnostics, de-provisioning platforms and re-provisioning affected services, or isolating network segments to prevent the spread of malware.
• Security enforcement: Enforcing automatic device and software load authentication; tracing identity, access, and trust mechanisms within and across corporate boundaries to provide secure services across firewalls.
• Logical isolation and privacy enforcement: Ensuring that a fault in a virtual platform does not propagate to another platform in the same physical machine, and that there are no data leaks across virtual platforms that could belong to different accounts.
• IT operations processes: Setting up generic micro IT operations as building blocks to standardize IT processes and enabling interoperability across heterogeneous system management products.

Figure 1: Service-oriented infrastructure framework
click image for larger view
 

As illustrated in Figure 1, the SOI provides an abstraction service layer to SOA applications. The backend IT infrastructure management also needs to be service-oriented in order to support and manage the applications and infrastructure. Service-oriented management (SOM) is a part of the service-oriented computing infrastructure that provides management services and interfaces to IT engineers. Platform management is delivered through a Standard Platform Interface (SPI), which is based on open standards, such as WS-Management.