Service Transition

1Introduction 2Serv. Mgmt. 3Principles 4Processes 5Activities 6Organization 7Consideration 8Implementation 9Issues AAppendeces

4. Service Transition Processes

4.1PLAN/SUPPORT 4.2CHANGE 4.3ASSET/CONFIG 4.4RELEASE/DEPLOY 4.5 VALIDATE/TEST 4.6EVALUATE 4.7KNOWLEDGE

4.4 Release and Deployment Management

Release and Deployment Management aims to build, test and deliver the capability to provide the services specified by Service Design and that will accomplish the stakeholders' requirements and deliver the intended objectives.

4.4.1 Purpose, Goal and Objective
The purpose of Release and Deployment Management is to:

The goal of Release and Deployment Management is to deploy releases into production and establish effective use of the service in order to deliver value to the customer and be able to handover to service operations.

The objective of Release and Deployment Management is to ensure that:

4.4.2 Scope
The scope of Release and Deployment Management includes the processes, systems and functions to package, build, test and deploy a release into production and establish the service specified in the Service Design package before final handover to service operations.

4.4.3 Value To Business
Effective Release and Deployment Management enables the service provider to add value to the business by:

Well-planned and implemented release and deployment will make a significant difference to an organization's service costs. A poorly designed release or deployment will, at best, force IT personnel to spend significant amounts of time troubleshooting problems and managing complexity. At worst, it can cripple the environment and degrade the live services.

4.4.4 Policies, Principles and Basic Concepts
4.4.4.1 Release Unit and Identification
A 'release unit' describes the portion of a service or IT infrastructure that is normally released together according to the organization's release policy. The unit may vary, depending on the type(s) or item(s) of service asset or service component such as software and hardware. Figure 4.15 gives a simplified example showing an IT service made up of systems and service assets, which are in turn made up of service components. The general aim is to decide the most appropriate release unit level for each service asset or component. An organization may, for example, decide that the release unit for business critical applications is the complete application in order to ensure that testing is comprehensive. The same organization may decide that a more appropriate release unit for a website is at the page level.

The following factors should be taken into account when deciding the appropriate level for release units:

Releases should be uniquely identified according to a scheme defined in the release policy as discussed in section 4.1.4.2. The release identification should include a reference to the CIs that it represents and a version number that will often have two or three parts, e.g. emergency fix releases: Payroll System v.1.1.1, v.1.1.2, v.1.1.3.

4.4.4.2 Release Design Options and Considerations
Service Design will define the approach to transitioning from the current service to the new or changed service or service offering. The SDP defines the service and solution design components to be transitioned to deliver the required service package(s) and service level package(s).

Figure 4.15 Simplified example of release units for an IT service
Figure 4.15 Simplified example of release units for an IT service

Common options for release and deployment that are considered in Service Design are discussed below. The selected option will have a significant impact on the release and deployment resources as well as the business outcomes. It is important to understand the patterns of business activity (PBA) and user profiles when planning and designing the releases.

'Big Bang' Vs Phased
Options for deploying new releases to multiple locations are illustrated in Figure 4.16 and described below:

Figure 4.16 Options for 'big bang' and phased rollout
Figure 4.16 Options for 'Big Bang' and Phased rollout

Figure 4.16 also illustrates a possible sequence of events over time as follows:

Variations of the phased approach include:

In the type of phased implementation illustrated above, it is only possible to employ this approach if the service has been designed to allow new and old versions to coexist. If this is not possible then the only alternative is to upgrade all affected parts together in a 'big bang' implementation. For elements such as documentation, for skilled staff this is rarely a problem; for many instances of hardware and software it is possible. For other transitions, such as those involving major network changes, it can be virtually impossible to achieve.

Figure 4.17 illustrates phased deployment to a number of different geographical locations. It assumes that new versions will work alongside the previous one. The example used assumes that new functionality is implemented first in the head office of the organization, then in a pilot branch and finally in the remaining branches. If there are a very large number of locations to deal with, it may still take a long time to implement the initial system or upgrades in all branches, thus increasing the likelihood of needing to support even more versions of the system in the live environment concurrently.

Push and Pull
A push approach is used where the service component is deployed from the centre and pushed out to the target locations. In terms of service deployment, delivering updated service components to all users - either in Big Bang or Phased form - constitutes 'push', since the new or changed service is delivered into the users' environment at a time not of their choosing.

A pull approach is used for software releases where the software is made available in a central location but users are free to pull the software down to their own location at a time of their choosing or when a user workstation restarts. The use of 'pull' updating a release over the internet has made this concept significantly more pervasive. A good example is virus signature updates, which are typically pulled down to update PCs and servers when it best suits the customer; however at times of extreme virus risk this may be overridden by a release that is pushed to all known users.

In order to deploy via 'push' approach, the data on all user locations must be available. Pull approaches do not rest so heavily on accurate configuration data and they can trigger an update to user records. This may be through new users appearing and requesting downloads or expected users not doing so, triggering investigation into their continued existence. As some users will never 'pull' a release it may be appropriate to allow a 'pull' within a specified time limit and if this is exceeded a push will be forced, e.g. for an anti-virus update.

Automation vs Manual
Whether by automation or other means, the mechanisms to release and deploy the correctly configured service components should be established in the release design phase and tested in the build and test stages. Automation will help to ensure repeatability and consistency. The time required to provide a well-designed and efficient automated mechanism may not always be available or viable. If a manual mechanism is used it is important to monitor and measure the impact of many repeated manual activities as they are likely to be inefficient and error-prone. Too many manual activities will slow down the release team and create resource or capacity issues that affect the service levels.

Figure 4.17 Phased deployment across geographical locations
Figure 4.17 Phased deployment across geographical locations

Many of the release and deployment activities are capable of a degree of automation. For example:

Designing Release and Release Packages
Figure 4.18 provides an example of how the architectural elements of a service may be changed from the current baseline to the new baseline with releases at each level. The architecture will be different in some organizations but is provided in this section to give a context for release and deployment activities. The release and deployment teams need to understand the relevant architecture in order to be able to plan, package, build and test a release to support the new or changed service. This helps to prioritize the release and deployment activities and manage dependencies, e.g. the technology infrastructure needs to be ready with operations staff ready to support it with new or changed procedures before an application is installed.

Figure 4.18 also shows how the service architectural elements depend on the Service Portfolio that defines the service offerings and service packages. Dependent services will need to be built and tested in Service Transition. For example an IT financial service may be dependent on several internal support services and an external service. For more details about the structure of services, see the Service Strategy and Service Design publications.

Figure 4.18 Architecture elements to be built and tested
Figure 4.18 Architecture elements to be built and tested

There are normally dependencies between particular versions of service components required for the service to operate. For example a new version of an application may require an upgrade to the operating system and one or other of these two changes could require a hardware change, e.g. a faster processor or more memory. In some cases, the release package may consist of a set of documentation and procedures. These could be deployed via a manual update or through an automatic publishing mechanism, e.g. to the SKMS/website.

Figure 4.19 Example of a release package
Figure 4.19 Example of a release package

A Release Package may be a single release unit or a structured set of release units such as the one shown in Figure 4.19.

The example in Figure 4.19 shows an application with its user documentation and a release unit for each technology platform. On the right there is the customer service asset that is supported by two supporting services - SSA for the infrastructure service and SSB for the application service. These release units will contain information about the service, its utilities and warranties and release documentation. Often there will be different ways of designing a release package and consideration needs to be given to establishing the most appropriate method for the identifiable circumstances, stakeholders and possibilities.

Where possible, release packages should be designed so that some release units can be removed if they cause issues in testing.

Valuable Release Windows

A UK government department is especially well placed to make full use of all available release windows. They work in a secure financial, low risk environment, with carefully planned changes scheduled well in advance and allocated to prearranged release windows, which are scheduled several months apart. Because of their careful and longer term planning, when a change proves unsuitable for release, i.e. tests are failed, alternative, quality-assured changes are usually available - prepared and tested but lower in business priority and so targeted at later releases. These can be accelerated to make use of the unexpected vacancy created by the test failure. The test and build process also allows elements of later scheduled releases to be slotted in for release, or successful components of the failed release to be implemented, even though the full products is not ready. This allows subsequent fuller release to be a 'smaller' product, therefore allowing further additional changes to be scheduled alongside them in later release windows.

Figure 4.20 Coordinating the deployment of service components
Figure 4.20 Coordinating the deployment of service components

Any significant new or changed service or service offering will require the deployment stage to consider the full range of elements comprising that service - infrastructure, hardware, software, applications, documentation, knowledge etc. Effectively this means the deployment will contain sub-deployments for elements comprising the service, as illustrated in Figure 4.20. The combination, relationship and interdependencies of these components will require careful and considered planning. Significant deployments will be complex projects in their own right.

To understand the deployment options a high level assessment of the deployment units, locations and environments may be required, for example:

4.4.4.3 Release and Deployment Models
A service may be deployed into the production environment in a number of ways. Service Design will select the most suitable release and deployment models that include the approach, mechanisms, processes, procedures and resources required to build and deploy the release on time and within budget.

The release methods during the early build and test stages may differ significantly from live operations so plan ahead to ensure that appropriate release methods are adopted at the right time.

Release and deployment models define:

Considerations in designing the release and deployment model include activities to:

4.4.5 Process Activities, Methods and Techniques
4.4.5.1 Planning
Release and deployment plans
Plans for release and deployment will be linked into the overall Service Transition plan and adopt the selected release and deployment model. The approach is to derive a sound set of guidelines for the release into production and subsequent deployment that can be scaled from small organizations to large multinationals. Although smaller organizations will have less complex environments, the disciplines detailed here are still relevant. Even within a single organization, the release and deployment plans need to be scalable since the extent of their scale of impact on the organization will vary, perhaps from impacting only one small specialist team in one location through to multinational impact on all users when introducing new desktop equipment and services, or transferring services to different suppliers.

Release and deployment plans should be authorized through Change Management. They should define the:

Pass/Fail Criteria
Service Transition is responsible for planning the pass/fail situations. At a minimum these should be defined for each authorization point through the release and deployment stage. It is important to publish these criteria to relevant stakeholders well in advance to set expectations correctly. An example of a pass situation before build and test is:

Examples of fail situations include:

Build and test prior to production
Build and test planning establishes the approach to building, testing and maintaining the controlled environments prior to production. The activities include:

Figure 4.21 provides an example of a model that can be used to represent the different configuration levels to be built and tested to deliver a service capability. The lefthand side represents the specification of the service requirements down to the detailed Service Design. The right-hand side focuses on the validation and test activities that are performed against the specifications defined on the left-hand side. At each stage on the left-hand side, there is direct involvement by the equivalent party on the right-hand side. It shows that service validation and acceptance test planning should start with the definition of the service requirements. For example, customers who sign off the agreed service requirements will also sign off the service Acceptance Criteria and test plan.

Figure 4.21 Service V-model to represent configuration levels and testing
Figure 4.21 Service V-model to represent configuration levels and testing

The V-model approach is traditionally associated with the waterfall lifecycle, but is, in fact, just as applicable to other lifecycles, including iterative lifecycles, such as prototyping, RAD approaches. Within each cycle of the iterative development, the V-model concepts of establishing acceptance requirements against the requirements and design can apply, with each iterative design being considered for the degree of integrity and competence that would justify release to the customer for trial and assessment.

LevelRequirements and DesignBuild/DeliverableValidation and Testing
Level 1
Customer/Business Needs
Structured definition of contract requirementsCustomer contract (based on Service Portfolio, SLP)Service Test and Evaluation
Determines whether a service can enable the users and customers to use the service to support their business needs (is fit for purpose and fit for use).
Level 2
Service Requirements
Service requirement specifications and SAC, traceable back to the contract requirementsService capability and resources to deliver against the SLA and service requirementsService Test
Test the Service Acceptance Criteria are met. Includes validation of service performance against the service level requirements and SLA in pilots, deployment and early life support.
Level 3
Service Solution
SDP, Service model, service environmentsSolution/system required to deliver service capability; includes the Service Management and Service Operations capabilitiesService Operational Readiness Test
To evaluate the integration and operation of the service capability and resources. It verifies that the target deployment organization and people are prepared to deploy and operate the new or changed service in the live environment, e.g. deployment systems and team, Service Operations, customers, users and other stakeholders. Tests include scenario-based testing such as simulation and service rehearsal.
Level 4
Service Release
 Release packageService Release Test
A test that the service components can be integrated correctly and that the release can be installed, built and tested in the target environments. Service release testing includes non-functional testing that can be performed at this level.
Level 5
Component and Assemblies
Component and assembly test specificationComponent or assembly of componentsComponent and Assembly Test
Test that a service component or assembly of components matches its detailed specification. Components or assemblies are tested in isolation, with a view to their delivering as specified, in terms of inputs generating expected outputs. Evidence of component quality or testing earlier in the chain may be obtained for test evidence, from both internal and external suppliers.
Table 4.8 Levels of configuration for build and testing

Further details on validation, testing and service evaluation are provided in sections 4.5 and 4.6. The test strategy defines the overall approach to validation and testing. It includes the organization of validation and testing activities and resources and can apply to the whole organization, a set of services or an individual service.

Typical levels of configuration for build and testing are shown in Table 4.8.

Various controlled environments will need to be built or made available for the different types and levels of testing as well as to support other transition activities such as training. Existing deployment processes and procedures can be used to build the controlled test environments. The environments will need to be secure to ensure there is no unauthorized access and that any segregation of duty requirements are met. The types of environments, both logical and physical, required during release and deployment include:

Planning Pilots
Pilots are useful for testing the service with a small part of the user base before rolling it out to the whole service community. It is important to determining the appropriate scope of a pilot (how much of the service is to be included in the pilot, size of department or user base). This is a key step in establishing the pilot effort. If the scope is too small then insufficient functionality and implementation variations will be trialled and the likelihood of significant errors not being discovered until full rollout is higher. If the scope is too large it will not deliver the speed and flexibility that deliver the benefits, but will effectively be a first rollout.

A pilot can be used to establish the viability of most, if not all, aspects of the service. But this will only happen if all stakeholders are actively involved in the pilot and use the service as it would be done in a full rollout.

The pilot should include steps to collect feedback on the effectiveness of the deployment plan. This can include:

Commitment to support the pilot is required from all involved parties. Obtaining that commitment can be a challenge since pilots typically will represent additional work for those users involved over and above their day jobs. Collaboration from suppliers and support staff (who may have to be supporting two versions of a service in parallel, or deliver a small separate unit dedicated to supporting the pilot) must also be obtained. Planning should accommodate rolling back a pilot before the full rollout of an authorized new service. New services tend to be piloted with test equipment and this needs to be rolled back to its original state. In addition, users who were part of the pilot should be working with the same components of a service as other users after the full rollout, not the setup put in place for the pilot. This simplifies, day-to-day operations in IT Service Management.

Although a pilot is often thought of as one trial in the production environment before rolling a service out across the full customer and user environment, there may be justification for a range of pilots, e.g. a pilot for deployment to each geographical region. Many considerations are relevant, with the best solution for a given circumstance being a balance between benefit and cost. Factors include:

Example of need for multiple piloting
A government organization delivers desktop IT services to all their staff - in corporate headquarters (HQ) and in locations throughout the world. When new or significant changes are to be rolled out, typically three parallel pilots are carried out to test the three levels of communication and support technology they have identified:
  • Those in HQ on direct network connection and with local dedicated support staff
  • Those in larger locations with reliable high-speed connection and semi-specialized local IT administrators
  • Those in smaller locations with unreliable communications and no trained local support.
Experience has shown that the three groups have different implementation and support issues and that the pilots in all three types of customer are worth the extra costs and complications.

Planning Release Packaging and Build
Planning the release packaging and build activities includes the activities to develop the mechanisms, plans or procedures for the following:

Deployment questionExamples
What needs to be deployed?Do you have a good understanding of the service and release that is being deployed? What are the components that make up the release package? What are the business drivers for the deployment? Is it required to meet a critical business need?
Who are the users?Which users are affected by the deployment? What language do they use? Do they need any special training?
Are there location dependencies?Are there any holidays, shut-downs or other interruptions to normal business at this location? What level of detail needs to be recorded, e.g. building, floor, room?
Where are the users?Are all the users and systems local to the deployment, or are some remote, and how will this affect the logistics?
Who else needs to be prepared well in advance?Do the service desk and support staff need training? Are there any access issues to be solved - security or physical?
When does the deployment need to be completed?Does the deployment need to be completed by a certain date and time or can it be completed by following a flexible schedule?
Why is the deployment happening?Is the deployment needed to fix a problem or is it required for some new functionality that has been requested, and do the users understand what is coming?
What are the critical success factors and exit criteria?How will you know that the deployment has been successful? Who will authorize the deployment? How will you know when the deployment is finished?
What is the current capability of the service provider?What are the current services, processes and Service Management capability - capacity, financial aspects, current systems and infrastructure?
Table 4.9 Questions to be answered when planning deployment

Deployment Planning
There are many planning considerations that need to be considered. Planners should be able to answer the questions included in Table 4.9.

Logistics and Delivery Planning
Once the overall deployment approach is understood, develop the logistics and delivery plans. These plans deal with aspects such as:

As well as the delivery aspects, there are typically consequential logistics to be dealt with, e.g. decommissioning and disposing of redundant items, including software and licences, hardware, skills, computer and staff accommodation, support contracts (utility supply, maintenance, cleaners etc.). There may also be a need for temporary equipment (e.g. swing equipment) or throwaway software that is required for the transition. If the transition plans call for any parallel running of services or equipment, this is particularly taxing from a logistics perspective, since double facilities are likely to be required for a short time. Once the logistics and delivery plans have been determined, they need to be communicated to all stakeholders, including formal notification to those consulted in deriving the plan.

Delivery is not sufficient; successful logistics requires that the components arrive and perform as required. Therefore deployment planning for all despatched items - hardware, software, documentation, and training - will address how components are tracked and documented on delivery. This should include:

Financial/Commercial Planning
Financial and commercial aspects will need to be specifically checked before the deployment and activities added to the deployment plans where necessary. For example:

4.4.5.2 Preparation for Build, Test and Deployment
Before authorizing the build and test stage, the Service Design and the release design must be validated against the requirement for the new or changed service offering. This should result in constructive feedback on the Service Design. Record, track and measure any risks and issues against the services, service assets and CIs within the service package, SLP, SDP or release package. Prioritize the issues and actions to ensure they can be resolved in a timely manner. Finally, produce a validation report and associated results ready for service evaluation.

An independent evaluation of the service and release design uses the validation report and results (see 4.6.5). This evaluation checks that the change to the services or service offering will deliver the predicted outcomes, i.e. the service expected by the user or customer. If there are issues, an interim evaluation report is prepared. This report lists the deviations from the SDP, a risk profile and recommendations for Change Management. If there are deviations in the service level requirements then the service package, SLP or SAC may be changed (via Change Management) and action should be taken to modify the proposed service release and related changes. Successful completion of the evaluation of the Service Design baseline ensures that service release build and test starts with a stable, baselined and approved design.

For some releases the Service Transition Manager may need to assign individuals or establish a team of competent people to execute the plans. If individuals are not dedicated there is risk that they may be diverted to work on other projects. Such risks need to be mitigated as they are often the cause of delays. On most occasions, the introduction of a technologyenabled service requires training for the release, deployment, build and test teams. The training needs of these groups will be at different levels. Recognition of the different skill sets, capabilities and competencies within the various groups is a useful prerequisite in identifying the necessary training. In specifying the training programme, the number of people that require training needs to be determined, and the way the knowledge can be provided needs to be considered. While the need for training differs from release to release, the impact of training can be significant. For example if support staff are spread around many locations, specific training, automated mechanisms, such as e-learning or computerbased training (CBT) solutions over the internet or intranet, may become an attractive proposition.

Examples of training needs include:

4.4.5.3 Build and Test
During the build and test stages, the common services and infrastructure need to be managed carefully since they can significantly affect the build and test of a technology enabled service and its underlying technology infrastructure. Key aspects that need to be managed during the activities to build and test a service or service offering are:

Configuration baselines of the controlled environments and the release package before and after an installation, build or deployment are recorded in the CMS to provide a restore point. The configuration information also needs to be updated to reflect the receipt and implementation of a release unit or the complete release package to a deployment group or target environment. The definitive version of the release package (approved in service release test) must be placed in the DML even where the release package consists only of documentation for a hardware upgrade. The release package must always be taken from the DML to deploy to the Service Operations readiness, service acceptance and live environments.

Release and Build Documentation
Procedures, templates and guidance should be used to enable the release team to take service assets and products from internal and external suppliers and build an integrated release package efficiently and effectively. Procedures and documents for purchasing, distributing, installing, moving and controlling assets and components that are relevant to acquiring, building and testing a release include:

'Proof of licence' is what a court will accept as proof of a legal entity having a licence. Each software manufacturer in general states the requirements for their proof of licence, so no hard and fast rules can be given here. As a general principle, proof of licence requires some form of evidence directly from the software manufacturer. There is a spectrum of types of evidence for having a proof of licence. Typical examples include:

The proposed solution should be documented to enable knowledge gathered during the build and test stages to be handed over to the Service Operations and Continual Service Improvement to be retained for future releases. It is important that the information is ordered and maintained in a systematic manner as during the build and test activities updates to the documentation will be required. The documentation includes:

Acquire and Test Input Configuration Items and Components
Configuration items and components (e.g. services, service assets) are acquired from projects, suppliers, partners and development groups. To prevent the acquisition of unknown and potentially risky components for a build it is essential to use CIs that have achieved a certain quality level or components from a catalogue of standard components that have been previously assessed, tested and authorized for use in specific conditions. Otherwise a change will need to be raised to assess the component and either incorporate it into the standards catalogue or accept it as a one-off exception for this release.

The acquisition activities include:

Verification activities to check the components destined for a release package or build include:

Issues, non-conformance, known errors and deviations reports about the quality of service components and any risks should be passed to the relevant stakeholders, e.g. quality assurance, CSI, Service Design.

Release Packaging
Build management procedures, methodologies, tools and checklists should be applied to ensure that the release package is built in a standard, controlled and reproducible way in line with the solution design defined in the Service Design Package. As a release package progresses towards production it may need to be rebuilt. For example: if a newer version of a CI or component needs to be incorporated quickly to fix errors; if the documentation needs to be updated.

The key activities to build a release package are:

If testing of a release package is successful, the release and the contents of the release package are placed under the control of Configuration Management, baselined and verified against the release design and release package definition. From this point all changes to the release package are managed through Change Management, e.g. to fix an error in testing. If at any step the testing of a release package does not complete successfully, reassessment and rescheduling of the release is managed through Change Management.

Build and Manage The Test Environments
Effective build and test environment management is essential to ensure that the builds and tests are executed in a repeatable and manageable manner. Inadequate control of these environments means that unplanned changes can compromise the testing activities and/or cause significant re-work. Dedicated build environments should be established for assembling and building the components for controlled test and deployment environments.

Preparation of the test environments includes building, changing or enhancing the test environments ready to receive the release.

An IT service is, on most occasions, built from a number of technology resources or management assets. In the build phase, these different blocks, often from different suppliers, are installed and configured together to create the solution as designed. Standardization facilitates the integration of the different building blocks to provide a working solution and service.

Automating the installation of systems and application software onto servers and workstations reduces the dependencies on people and streamlines the procedures. Depending on the release and deployment plans, the installation may be performed in advance (for example, if equipment is being replaced) or it may have to occur in situ in the live environment.

The physical infrastructure elements, together with the environment in which they will operate, need to be tested appropriately. Part of the testing may be to test the replication of the infrastructure solution from one environment to another. This gives a better guarantee that the rollout to the production environment will be successful.

Test environments must be actively maintained and protected using Service Management best practices. For any significant change to a service, the question should be asked (as it is for the continued relevance of continuity and capacity plans): 'If this change goes ahead, will there need to be a consequential change to the test data?' During the build and test activities, operations and support teams need to be kept fully informed and involved as the solution is built to facilitate a structured transfer from the project to the operations team.

4.4.5.4 Service Testing and Pilots
The testing activities are coordinated through test management, which plans and controls the testing execution that is described in section 4.5. Testing aims to build confidence in the service capability prior to final acceptance during pilot or early life support. It will be based on the test strategy and model for the service being changed.

The test criteria reflect the anticipated conditions in which the service is expected to operate and deliver benefit. However, these surrounding circumstances may change, and in many modern situations such change is almost inevitable and often unpredictable. These changes and their impact on service testing and acceptance must be observed, understood and documented. Their consequences need to be expressed in terms of changed Acceptance Criteria and updates to the service package, including the SLP. This will need the collaboration and input of the business, customers and other affected stakeholders, which may well include suppliers and operations. The Service Designer will be involved in making any amendments since this knowledge may assist in building in additional and relevant flexibility to designs of future new or changed services.

An example of tests that can be executed during release and deployment is shown in Figure 4.22. Further details of these tests are described in section 4.5 on validation and testing. In practice, the test types overlap the different levels of testing to provide a full range of testing across the service life.

A service release test checks that the service components can be integrated correctly and that the release can be installed, built and tested in the target environment.

Service Operations readiness testing ensures that a service and its underlying application and technology infrastructure can be transferred into the production environment in a controlled manner. It provides a level of confidence that the new or changed service will provide the level of service specified in the service requirements and service level requirements. However, it is too early to finalize the SLA at this point. The SLA is finalized in the pilot or more usually in early life support before the Service Transition is closed. The service operational readiness test aims to:

Figure 4.22 Example of service testing through Service Transition
Figure 4.22 Example of service testing through Service Transition

Tests that are conducted as part of service operational readiness test include:

Service Rehearsals
One testing method is to simulate as much of the service as possible in a service rehearsal (sometimes referred to as 'model office'). A service rehearsal is a simulation of as much of the service as possible in an extensive and widely participatory practice session. It is the ultimate stage of internal testing, the last stage before any public live running. This is like a 'dress rehearsal' of a play, setting out all the elements - costume, lighting etc. - in a last private run-through of the performance. It can deliver significant benefits by establishing errors and unworkable procedures before they impact the business in live operation. However, they are complex, time consuming and relatively expensive to prepare, deliver and document. A careful and deliberate balance is therefore required between the anticipated costs and the risk damage profile that they could prevent.

A service rehearsal takes place just before deployment of the service; if held too early there is a significant chance that the environment, technology, people and legislation into which the service is being released will change and invalidate the results. If too close to the declared release date any issues found will not be addressed before the service goes live.

The objectives of the service rehearsal include:

The service rehearsal requires adequate representation from all stakeholders, with commitment to providing staff for - typically - a full day rehearsal for a new or significantly changed service. It is often beneficial to involve 'ordinary' representatives of the stakeholder community, not those with previous experience or knowledge of the service. Typical mistakes will be more likely to come from typical users - those who have been involved in design and development will find it impossible to 'unlearn' and will be coloured by their expectations of service behaviour.

The focus of a service rehearsal is typically on one day of actual rehearsal, but successful delivery of a service rehearsal involves more stages, including preparation and analysis, mirroring the Plan-Do-Check-Act cycle. Typical stages for a service rehearsal would include the following activities.

Plan - prepare for the day
Request for a service rehearsal - the project or service implementation teams consider that a service rehearsal would be appropriate and trigger the process with a request.

Tasks include the following:

Do - deliver the rehearsal
Hold meetings to:

Check - document the day
Tasks include:

Act - take action following the rehearsal
Considering the results from the rehearsal, the options will be:

Pilots
Pursuing the theatrical analogy seen in service rehearsal, if the service rehearsal is the 'dress rehearsal' - the last practice before being seen by the public - then the pilot is the 'off Broadway' run of a play. It is done for real and in public, but for a small audience only and with the expectation of further (hopefully minor) polishing of the performance, script, scenery and effects. Conducting a pilot is easier to control as it is deployed to a smaller environment/user base.

A pilot sets out to detect if any elements of the service do not deliver as required and to identify gaps/issues in Service Management that put the service and/or the customer's business and assets at risk. It does not need to cover all service and system functionality, but will focus on the areas of risk and perform enough of the service to determine if it will work sufficiently well in deployment. It aims to ensure that the service capability supports delivery of the service requirements and service level requirements. As far as possible it should check that the service utilities are fit for purpose and the warranties are fit for use.

Establish clear objectives for the pilot implementation such as:

As there are likely to be design changes and improvements that need to be built into the release before full deployment, it is important to agree how these will be funded up front. It is also important to ensure that there is a common understanding about how the pilot implementation will be signed off.

During the pilot the release and deployment team should:

When the release has been in use for a sufficient period during a pilot it is important to check that the service is capable of delivering the requirements of the customer, user and the Service Design as well as the predicted outcomes (although not all these will be realized at this point).

If the pilot is of sufficient length, it may be appropriate to conduct an independent evaluation to compare the actual vs predicted service capability and performance (specified in the Service Design) on behalf of the stakeholders, users and customers. This evaluation includes a risk assessment on whether the service will continue to deliver the service requirements, e.g. service levels and warranties.

The outputs from a successfully delivered service pilot will include:

4.4.5.5 Plan and Prepare For Deployment
The planning and preparation activities prepare the deployment group for deployment. This is an opportunity to prepare the organization and people for organizational change; see section 5.2. The overall approach to planning the deployment is described in release and deployment planning (see paragraph 4.4.5.1). During the actual deployment stage the detailed implementation plan is developed. This includes assigning individuals to specific activities. For example a specific individual may be assigned to deliver training for a training activity on the deployment plan.

The entry criteria for planning and preparing a target deployment group or environment include:

An example of the deployment activities that apply to the deployment for a target group is shown in Figure 4.23.

Figure 4.23 Example of a set of deployment activities
Figure 4.23 Example of a set of deployment activities

Preparing for deployment includes assessing each deployment group's readiness to receive and implement a release package, identifying gaps that need to be filled and planning the activities required to deploy, transfer or decommission/retire services or service assets. It will also include transferring a service or a service unit as well as move and disposal activities.

Assessment
Although the deployment assessment should be conducted early, it should be revisited periodically. The results of this assessment are fed into detailed implementation planning for the target deployment group.

The readiness assessment for a deployment group identifies:

The aspects to assess include:

Develop Plans and Prepare For Deployment
Planning for a specific deployment includes assigning specific resources to perform deployment and early life support activities. While developing these plans, identify and assess risks specific to this deployment group by using the service model to identify business and service critical assets that have the highest risk of causing disruption. The activities include:

The next step is to verify the detailed deployment plans, perform any deployment readiness tests and raise an RFC to be authorized through the Change Management process. The service is then ready for deployment.

4.4.5.6 Perform Transfer, Deployment and Retirement
The following activities provide an example of the different aspects that will be performed in the order specified on the deployment plan.

Transfer Financial Assets
Changes and transfers of financial assets need to be completed as part of deployment. This will include but is not constrained by the following:

Transfer/transition Business and Organization
Transfer of a business unit, service or service unit will involve change to the organization itself. The subject of organizational change is addressed in Chapter 5. Activities that need to be performed include:

When the change includes a transfer of service provider, e.g. new outsourcing, insourcing, change of outsourced provider, then some specific elements need to be considered, e.g. organizational change, quick wins to avoid confusion and higher staff turnaround.

Competent people with the right skills are required to perform the deployment, operate and manage the new or changed service in the business, customer and service provider organization. The related activities include:

Deploy Processes and Materials
Deploy or publish the processes and materials ready for people involved in the business and service organization change, e.g. users and Service Operations teams that need to execute the new or changed processes. The materials may include policies, processes, procedures, manuals, overviews, training products, organizational change products etc.

Training people to use new processes and procedures can take time, particularly for a global deployment to thousands of people.

Deploy Service Management Capability
Deploy new or changed processes, systems and tools to the service provider teams responsible for Service Management activities. Check that everyone is competent and confident to operate, maintain and manage the service in accordance with the service model and processes. Remove or archive redundant services and assets, e.g. processes, procedures and tools.

During deployment monitor the service against the service model and performance standards as far as possible.

Transfer Service
Transferring a service will also involve organizational change described earlier in this section. The issues around transferring a service and the activities that need to be performed include:

When the change includes a transfer of service provider, e.g. new outsourcing, insourcing, change of outsourced provider, then some specific elements need to be considered that include:

Deploy Service
Deploy the service release and carry out the activities to distribute and install the service, supporting services, applications, data, information, infrastructure and facilities. These will include:

Decommissioning and Service Retirement
Some specific aspects need to be considered for decommissioning and retiring services and service assets. For example the procedures for retiring, transferring (e.g. to another budget holder) or redeploying service assets need to take into account any security, confidentiality, licensing, environmental or other contractual requirements. This includes:

Remove Redundant Assets
A comprehensive understanding of the assets used by a retired service needs to be gained and managed. With a full understanding any redundant assets can be identified and removed, therefore potentially saving licence fees, liberating capacity and preventing accidental use. Failure to develop and properly perform these activities can result in:

As part of the clean-up activities it is important to delete or archive redundant data, information and records related to the previous service or products. The full scope and scale of a service or service asset needs to be considered, and this should extend to the following areas:

4.4.5.7 Verify Deployment
When the deployment activities are complete, it is important to verify that users, Service Operations, other staff and stakeholders are capable of using or operating the service. The tests should specifically verify that:

This is a good point to gather feedback on the deployment process to feed into future improvements, e.g. using satisfaction surveys.

Report any issues and incidents and take corrective actions as necessary. Successful confirmation of the deployment verification triggers the initiation and launch of early life support for the deployment group.

4.4.5.8 Early Life Support
Early life support (ELS) provides the opportunity to transition the new or changed service to Service Operations in a controlled manner and establish the new service capability and resources. An example of the ELS activities is shown in Figure 4.24.

Figure 4.24 Example of early life support activities
Figure 4.24 Example of early life support activities

In Service Design, the stakeholders will have agreed the entry and exit criteria from early life support but it may be necessary to finalize the performance targets and exit criteria early in this stage. This can help to understand the deployment verification process and set customer and stakeholder expectations about the handover of the service to Service Operations.

ELS provides appropriate resources to resolve operational and support issues quickly, centrally and locally, to ensure that the users can use the service to support their business activities without unwarranted disruption. The deployment teams should analyse where users and support resources will experience issues and problems, perhaps based on previous experience; for example, clarification about:

During ELS, the deployment team implements improvements and resolves problems that help to stabilize the service. The Continual Service Improvement publication provides relevant information on measurement and service improvements. The deployment resources will gradually back out from providing the additional support as the users and service teams become familiar with the changes and the incidents and risks reduce.

Figure 4.25 Illustration of the benefits of targeted early life support
Figure 4.25 Illustration of the benefits of targeted early life support

Metrics for the target deployment group or environment measure service performance, performance of the Service Management and operations processes and teams and the number of incidents and problems by type. The deployment team's aim is to stabilize the service for the target deployment group or environment as quickly and effectively as possible. An example of a deployment performance graph is shown in Figure 4.25.

Variation in performance between different deployment groups and service units should be analysed and lessons learned from one deployment used to improve subsequent deployments.

The example shown in Figure 4.25 shows the number of incidents for two branches of a retail organization that have the same number of users and the same deployment schedule. In deployment A the incident levels have reduced faster. On further investigation the Service Transition manager discovered that the team responsible for Deployment A was more competent at training users and transferring knowledge to the service desk so that they could help users to be more effective more quickly.

During ELS, the deployment team should ensure that the documentation and knowledge base are updated with additional diagnostics, known errors, workarounds and frequently asked questions. The team should also resolve any knowledge transfer or training gaps.

At agreed milestones in early life support, it is important to assess the issues and risks, particularly those that impact the handover schedule and costs. Service Transition monitors the performance of the new or changed service in early life support until the exit criteria are achieved. These include when:

4.4.5.9 Review and Close A Deployment
When reviewing a deployment the following activities should be included:

Each deployment should consider whether any relevant issues have been detected that should be passed through to CSI, such as:

Deployment is completed with a handover of the support for the deployment group or target environment to Service Operations.

A post implementation review of a deployment is conducted through Change Management.

4.4.5.10 Review and Close Service Transition
In order to finalize that a Service Transition is completed, there should be a formal review carried out that is appropriate to the scale and magnitude of the change. A review of the Service Transition should include:

Independent evaluation of the service release uses the outputs from deployment. This evaluation checks the actual performance and outcomes of the new or changed service against the predicted performance and outcomes, i.e. the service expected by the user or customer. An evaluation report (see 4.6.6) is prepared that lists the deviations from the SP/SLP/SDP, a risk profile and recommendations for Change Management. If there are deviations in the service level requirements then the service package, SLP or SAC may need to change (via Change Management, in agreement with the customer representative and other stakeholders). Successful completion of the evaluation ensures that the service can be formally closed and handed over to Service Operations and CSI.

A transition report should be produced that summarizes the outcomes. As part of producing such a report a post transition workshop could be held involving all parties as a 'lessons learned' exercise. Lessons learned and improvements are fed into Change Management for a post implementation review and into Continual Service Improvement for future transitions.

4.4.6 Triggers, Input and Output, and Inter-process Interfaces
The release process commences with receipt of an approved RFC to deploy a production-ready release package. Deployment, commences with receipt of an approved RFC to deploy a release package to a target deployment group or environment, e.g. business unit, customer group and/or service unit.

The inputs are:

The outputs are:

Deployment is completed with a handover of the new or changed service to operations on successful completion of the post implementation review of the deployment conducted within Change Management.

4.4.7 Information Management
Throughout the deployment process, appropriate records will be created and maintained. As configuration items are successfully deployed, the CMS will be updated with information such as:

Other data and information will also be captured and recorded within the broader service knowledge management system. This could include:

Deployment information, history of the deployment itself, who was involved, timings etc.

As part of the clean-up activities it is important to delete or archive redundant records related to the previous service or products.

4.4.8 Key Performance Indicators and Metrics
4.4.8.1 Customers Or Business
Indicators include:

4.4.8.2 Service Providers
Indicators include:

4.4.9 Challenges, Critical Success Factors and Risks
Challenges for release and deployment include:

Critical success factors include:

Risks to successful release and deployment include:

Supporting Material
  1. META - Release Mgmt
  2. Becta- Intro to Release Management
  3. MOF - Stabilize Service Management Function
  4. MOF - Deploy Service Management Function

[To top of Page]


Visit my web site