Real Time Infrastructure Technologies: Shifting the Boundaries of Service Management

Shifting Boundaries

In the previous article, we looked at the arrival of real-time infrastructure (RTI) technologies — systems that can self-configure, self-manage, and self-optimise — such as those being provided by Sun, HP, and IBM.

We looked at how RTI will change the rules of service management: the time-consuming tasks of infrastructure management will be automated, enabling the infrastructure to become more dynamic and adaptable. In this article, we start to explore the implications for ITIL.

Real-time infrastructure (RTI) challenges some of the long-standing organisational and cultural rules surrounding the use of IT. The necessary changes will not suit every company, but are critical if RTI is to be deployed to ensure business success.

Change creates opportunities that can be capitalised upon by those who can adapt quickly. A successful RTI transition needs to consider the organisation's ability to accept change and its willingness to realign the traditional IT and business roles to accommodate both the technology changes and, more importantly, the efficiencies that can be gained by the business process changes.

For example, many companies buy IT on a project-by- project basis, but the true benefits of RTI can only be realised if the purchasing model changes to enable cross line of business procurement. Once all the IT resources are virtualised and assigned under an RTI model, then ownership is immaterial. It is therefore essential that any new RTI model contain a strong charge-back facility to enable the actual cost of the service to the business to be identified.

The focus today is often on reducing the cost of IT; however, what we should really be concerned about is the value of the service to the business and the costs of running the service as relative to that value.

Currently, many IT projects never see the light of day due to the imbalance of implementation and the decommissioning costs for even a short-term project. As IT organisations adopt RTI they will be able to implement, develop and test new IT services faster, enabling short-term projects to become economically viable.

Implementation costs will be minimal, as current assets can be assigned to the project when there is a trough in usage. In addition, there would be few decommissioning costs as the assets can be reassigned to the resource pool, only being decommissioned at end of life.

Many IT departments base their operational models on one of the leading industry standards; ITIL — service delivery and service support model, CobIT, BS15000 and others. These frameworks provide excellent guidance for running operational environments, but unless they evolve and keep pace with technological change they will become redundant.

The rest of this article and subsequent articles examine in detail the impact of the new RTI models on ITIL-based service management, specifically in the area of service delivery.

Service Delivery and Support — Time For Change?

In the ITIL model, service delivery consists of service-level management, IT financial management, capacity management, service-continuity management and availability management — while service support today consists of service desk, incident management, problem management, change management, release management and configuration management.

With the advent of RTI, even this definition becomes blurred; however, to aid clarity we will use these headings as a basis for exploring the expected significant differences RTI will bring. Here, we look in detail at the service delivery disciplines.

Service Level Management

In current best-practice terms, service-level management is the term for the process of planning, coordinating, drafting, agreeing, monitoring, and reporting on service-level agreements. In addition to the ongoing review of service achievements to ensure that the required service quality is maintained and improved upon, service level agreements provide the basis for managing the relationship between the IT service provider and the business.

In today's infrastructures, service level management is necessary to ensure that we can determine the level of IT service needed to support the business. The service-level objectives are then translated into the respective services and then monitored to identify whether the required service levels are being achieved.

Currently, service level management is seen as necessary to the delivery of today's services. But in a fully automated RTI environment, service level management becomes essential; it is the key to translating the business requirements.

As an example of RTI, Sun's N1 fabric operating environment takes service-level objectives and business goals and translates them into operational parameters. Because the service level objectives form an integral part of the RTI model, this automation results in greater agility.

The fundamental difference with RTI with regard to service level management is that it is a function that can no longer be ignored; without these operational parameters the service cannot exist.

To gain the maximum benefits from RTI, companies will need to fully understand the services they are delivering to the business. In many companies, the preparation required to enable them to implement RTI will help to accelerate their process maturity level within the service level management function.

In addition, once RTI is running, if it cannot meet service levels, the operating environment can alert the IT service-level architects to modify the service description. It may be that either the service levels are incorrect, or the way that the service has been defined to the RTI environment may need to be changed.

This represents a vast improvement on the levels of automation that are currently available for monitoring and reporting upon service level management. This in turn facilitates a transition from periodic (weekly, monthly) service-level reporting to true real-time responsiveness. There will be a more intuitive alignment between management decisions and business requirements through RTI's focus on managing the whole service.

The service level manager will become pivotal in the IT/business relationship because defining a service description and assigning service priorities will actually dictate the way the service is delivered. As the demand for a service increases or decreases due to business requirements then RTI will adjust to the change automatically.

The key challenge facing many businesses will be translating their current loosely written service levels into meaningful service definitions. IT staffs will need to truly understand and be aligned to the goals of the business to enable them to design and implement the service definitions.

IT Financial Management

Whilst almost every business would agree that effective financial management is essential, many do not truly identify and track the real cost of IT, let alone its value. Effective IT financial management clarifies the true cost of service and makes it possible for the customer(s) to judge whether value for money is being provided.

RTI integrates the accounting into the service definition and makes it part of the service lifecycle.

It provides monitoring services at runtime, which enables RTI to track resource usage. It then collects, aggregates and correlates this information according to the specific requirements.

Most companies are looking at total cost of ownership (TCO) and trying to reduce IT spending. Whilst it is important to ensure that money is spent effectively, cost reduction in cost does not necessarily mean a more efficient service. We all recognise the value of TCO; however, the real issue is being able to determine the value of the service to the business. Support costs should be measured as a percentage of the value of the service to the business.

RTI implementations such as N1 provide accurate and detailed total cost of ownership figures, which are based on actual usage. These can then also be used in conjunction with the service definitions and business requirements to identify the true value of the service to the business.

This in turn will lead to better prioritisation in capital operating expenditures, including the most effective way of reducing costs without detracting from overall business value.

As a result, there is a significant change in the risk management associated with major projects; risk management is now more about meeting the real need effectively and in time, rather than the issues associated with major capital expenditure.

Capacity Management

The goal of capacity management process is to ensure that cost justifiable IT capacity always exists and that it is matched to the current and future identified needs of the business. It is therefore imperative that the capacity management function understands the business requirements, the operational structure, and the IT Infrastructure.

• Resource capacity management
• Service capacity management
• Business capacity management

RTI enables you to automate all three in a new way. It can remove the laborious requirement to collate information and interpret usage trends; these are an intrinsic part of the automation. For example, Sun's N1 allows peaks and troughs to be catered to in the service definition, which ensures the resource is available.

N1 will also alert you when the total capacity is expected to be exceeded. At this point, it is possible to add more resources to the pool or even to lease capacity for a temporary peak from another virtual resource pool external to your company.

RTI is the perfect vehicle for capacity on demand; through the use of the service definitions, it can ensure that the IT services continue to satisfy changing business expectations. Individual service peaks and troughs can be catered to when defining the service descriptions, ensuring the optimum utilisation of all elements of capacity at the service level.

Correctly approached, this is seen by many as a key area for RTI: the automation of capacity management reduces the need for costly spare capacity in the IT infrastructure. But, again, this can only be successfully achieved by ensuring rigorous processes linked to business requirements.

IT Service Continuity Planning

Most businesses are highly dependent on IT facilities for their day-to-day operations. Therefore, management should take the necessary steps to ensure continuity of processing, both through reducing the risks to processing and by making contingency plans that enable business operations to carry on in the event of disruption to processing.

As part of planning Business Continuity arrangements, organisations should consider the value produced by business operations, and the cost of not having them. Apart from straightforward revenue, there will also be considerations of image and reputation, potential loss of future business in commercial organisations and possible non-compliance with contractual and other legal obligations.

RTI can leverage a large virtual pool of resources that will respond to failures in any of the components that are managed in a transparent way, according to predefined rules set by the business. This should eliminate the disruption of temporary unavailability of a component and could significantly reduce the impact of even a whole site losing operational capability.

Managed correctly, a single-site disaster would appear as no more than a temporary reallocation of resources around the virtualised resource pool, with little or no effect on the services being offered.

For those companies not wishing to have that much spare capacity, it is likely that the service provider of the future will take customer service definitions as defined for an RTI infrastructure and overlay them onto their existing virtualised resource pool with little difficultly.

Either way, the consideration is less about "disaster recovery," and more about the how the policies applied to the RTI infrastructure enable continuity of service.

Availability Management

Availability Management is a set of processes that ensure that the availability of IT services is provided economically during their life cycle. It optimises the availability of IT services, to ensure that customers' cost-justifiable requirements are met.

A service can only be regarded as available if business areas are able to carry out all functions when needed to support business operations. It measures services critical to a business area and will be driven by that area. It will reflect the impact of failure upon business operations. Availability will be incorporated into service level agreements and their underpinning operational level agreements and contracts.

As an example of RTI, Sun's N1 is designed to enhance the existing highly available and resilient infrastructures. It leverages a large virtual pool of resources, ensuring that services can be configured and reconfigured at short notice to meet the demands of the business — so-called "cable once, provision forever."

Availability management as a discrete discipline will cease to exist in an environment where resources can be pooled and shared to such an extent.

Rather, it becomes an intrinsic specification within the service level and capacity requirements definitions. But, again, clear definition of the policies is essential for a successful RTI implementation.

Overall Impact on Service Delivery

Service delivery functions today are quite often the last ones implemented; however, in the future these will be the activities that must be considered at service inception and designed into the infrastructure and the applications that it supports.

Service level management becomes a crucial function for the definition of services without which the service cannot run. This is a key link with business requirements and policies; service level objectives must be defined before a service can go live.

IT financial management becomes a mature integrated function within the service lifecycle, allowing true costs of usage to be identified. This will enable IT departments to demonstrate the cost effective services they are running. Also, the risk management of finance now has different priorities.

IT service continuity management advances should eliminate the need for complex standby arrangements. Many customers will be able to utilise there own assets in other data centres or other companies in a much more transparent way than is possible today.

Capacity Management as a discipline becomes almost non-existent as the RTI framework optimises all IT assets all the time. Capacity monitoring also frees up IT resources during a decommissioning stage, allowing almost instant reuse of assets.

Availability management will cease to be a discrete discipline in an environment where resources can be pooled and shared to such an extent.

These last two are most impacted by the automation on service delivery — the move from manual and reporting activities to encoded procedures, which are interpreted by the RTI technology.

But the first three (service level management, IT financial management and capacity management) become very significant. They are the interpretation of business policies into the IT service management framework, and so the need for well-defined, mature processes becomes imperative to enable the automation. The consequence of this will be clearer alignment between business needs and IT delivery.

In the next article, we shall explore the impact of RTI on Service Support.