Shared Service Data Centre Final Options Analysis
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
2
© PTS Consulting 2010
1
EXECUTIVE SUMMARY
1.1
PROJECT INTRODUCTION This document concludes the second stage of the Shared Service Data Centre (SSDC) project, undertaken by PTS Consulting, reporting on the findings of the project. The Shared Service Data Centre project aims to develop and provide a cost effective data centre service to be shared, initially amongst three Higher Education Institutions (HEI), working together in partnership. It is envisaged that a modern, purpose-built facility will reduce risk in addition to reducing long-term costs, and will be delivered within the unique time constraints of each institution. The aims of this stage of the project were to identify options for: The Customers and HEI Partners – determining what the HEI partners’ business needs and desires are and how they could be met by a shared data centre service. Discovering who else may want to make use of the service (customers), and therefore what it should offer The data centre service model – discovering the services that service provider partners would be interested in offering. Determining what the future service might consist of given the service provider partner appetite, and the needs of the HEI partners and customers. Researching the competing services, either currently available or in development The shared partnership model – understanding how the HEI and service provider partners would prefer to work together. Discovering what may influence how they work together, and determining which Partnership model options are available
1.2
FINDINGS The project team determined that the customers and HEI partners have an immediate need for efficient, cost effective data centre services covering key services such as basic colocation (data centre power and space); disaster recovery and service continuity services; through to more advanced IT infrastructure services such as on-demand virtualised server and storage services. The project team discussed these initial requirements with identified service provider partners to determine the types of services that could be offered in support of the HEI needs. There was clear enthusiasm for the project, with service provider partners offering to provide infrastructure as a utility service1. This could greatly assist the HEI community in reducing the need to provide large up-front capital investments for IT equipment.
1 Utility Services are where the service provider provides its services or infrastructure platform without any initial reward, with the customer paying only for the services they use. SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
3
© PTS Consulting 2010
In exploring the shared service partnership model options for a shared service data centre, it was determined that a choice of 3 options were appropriate: 1. Consortium - the HEI partners would procure services from service provider partners and provide back (broker) services to the HEI partners and invited customers 2. Joint Venture - the HEI partners, under a legal entity, engage with a private partner who invests finance into the joint venture to set up the service. Both parties (the HEI partners legal entity and the private partner legal entity) gain from the rewards of the service 3. Technical Sourcing Authority - the HEI partners pool their service requirements and one of the HEI partners agrees to procure the combined services for the 3 HEI partners thereby attracting better pricing through economies of scale
1.3
CONCLUSION At the concluding stage of this project, as the studied elements were linked together and a basic shared service data centre assessed for the initial 3 HEI partners, it was concluded that the model is financially viable only if greater virtualisation can be achieved than has currently been (-£10m). However, without aggressive virtualisation (i.e. staying at current levels) it is financially beneficial for each institution to upgrade its own facilities rather than a shared facility (-£4.8m). This result is due to a combination of the physical space requirements (number of cabinets), power costs, and new build investment costs. If the institutions alternatively chose to use existing 3rd party supplied collocation space, the costs are much higher (+£27m) than upgrading existing institution facilities in a non-shared environment. However, the costs modelled do not take into account all factors such as alignment in migration times. The shared model would work, however, in the context of a much larger shared service targeted at the wider HE community. This is because greater economies of scale and purchasing power would result in greater savings. The present requirements from the existing HEI partners, taking into consideration efficiencies in configuration of existing infrastructure and implementation of virtualisation projects, could be shrunk from an existing data centre capacity of 70 racks of IT equipment, to less than 45 racks. When the HEI partner’s requirements were mapped on a timeline it was also highlighted that the 3 HEI partner’s data centre service requirements were not aligned, with Derby requiring service immediately, Leicester requiring services within 2 years, and Salford requiring services in 3 years. The consequence of this timing is that the SSDC is a long-term project, to the extent that it carries a high threshold cost (as building a new data centre would), and therefore requires substantial upfront investment which will not pay back for a number of years. Whilst this weighs in favour of a much larger customer group, a larger group would require even greater initial capital investment. This higher risk would be rewarded in higher longer-term savings. Taking into account the size of the 3 HEI’s IT requirements it was agreed by the SSDC Project Steering Group that this would not alone attract enthusiasm from the service provider partner community. A larger group of HEIs, and therefore a much larger opportunity, would. It was also determined by the Projects Steering Group that this service would be better purchased from a data centre service brokerage organisation. The project steering group determined the assumption that the shared service data centre would not be managed or staffed by institution personnel. The purpose of this was twofold, 1) to
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
4
© PTS Consulting 2010
repurpose institution staff to support their businesses, 2) for the institutions to benefit from expertise and industry best practices provided by the service provider community. It was therefore concluded that the basis for setting up a 3 HEI partner shared service data centre was no longer viable, primarily due to the lack of enthusiasm from the service provider community given the scale of the 3 HEIs IT requirements. The project steering group did, however, determine that there is a requirement for some form of national data centre service providing data centre services to the wider education community.
1.4
RECOMMENDATIONS It was determined that a national data centre service scope of services should not be limited to the supply and support of data centre services, but also include a number of other supporting services which include: Change Consultancy Service – assisting HEIs in winning the case for change within their organisations, identifying, planning and migrating services to the shared service data centre Service Management – managing the service provider partner community to provide optimum levels of service, and developing services in-line with defined institution business strategies. Additionally, setting up an HEI customer relationship management process to assist HEIs in managing their services and working with the HEI community to promote the services Subject Matter Expertise – in providing access to subject matter experts in the area of delivering technology solutions to the HEI community. These would include, server and storage technology, IT service management, data centre facility management, IT security etc. Data Centre Sourcing – a national service would acquire appropriate data centre space (must meet minimum levels of service and build) as required to meet future capacity requirements. Data centre space would be expected to be centrally procured through a European procurement framework, and then provided to the HEI customers as part of a full data centre service solution. The NEDCS would provide a 24x7 support service for all services provided, these may be either developed internally or procured from a service provider partner
1.5
NEXT STEPS Upon agreement that a national service is the most appropriate path to take the SSDC project forward, it is recommended that an incubator or pilot project be undertaken. This project would focus upon testing the assumptions made here and the rationale for such a service, by putting in place its basic elements, procuring data centre services, and transitioning the services of a small number of pilot customers. In developing this incubator project the services could be developed from procured service provider partners and consultancies. Should the Incubator Project meet delivery and service targets, it could be further developed into the national service providing the full extent of required services. In proceeding through a controlled test environment first, the results can be communicated back into the community and would quell any concerns that the community may have for such a service. It is therefore recommended that a business case be drawn up for the next stage of the
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
5
© PTS Consulting 2010
project to build an Incubator Project which, if successful in meeting key success criteria, would be developed into the full national service.
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
6
© PTS Consulting 2010
Shared Service Data Centre Final Options Table of Contents Section
Page Number
1
EXECUTIVE SUMMARY ......................................................................................................................... 3
1.1
PROJECT INTRODUCTION .................................................................................................................................3
1.2 1.3
FINDINGS..........................................................................................................................................................3 CONCLUSION....................................................................................................................................................4
1.4 1.5
RECOMMENDATIONS ......................................................................................................................................5 NEXT STEPS ......................................................................................................................................................5
2
DOCUMENT CONTROL ......................................................................................................................... 9
2.1 2.2 2.3
DOCUMENT AUTHORITY ..................................................................................................................................9 PROJECT INFORMATION ..................................................................................................................................9 CLIENT DISTRIBUTION ......................................................................................................................................9
2.4
DOCUMENT HISTORY .......................................................................................................................................9
3
PURPOSE OF THIS DOCUMENT ........................................................................................................... 10
4
PROJECT DEFINITION ......................................................................................................................... 11
4.1
PROJECT BACKGROUND ................................................................................................................................ 11
4.2
PROJECT BRIEF .............................................................................................................................................. 12
4.3 4.4
PROJECT OBJECTIVES .................................................................................................................................... 12 PROJECT APPROACH ..................................................................................................................................... 13
5
CUSTOMER REQUIREMENTS .............................................................................................................. 14
5.1 5.2 5.3 5.4
HEI PARTNER REQUIREMENTS ...................................................................................................................... 14 POTENTIAL CUSTOMERS ............................................................................................................................... 15 REQUIREMENTS IDENTIFIED FROM SURVEY OF HEI AND FE COMMUNITIES ............................................... 15 TRANSITION REQUIREMENTS ....................................................................................................................... 16
6
SERVICE PROVIDER PARTNERS ........................................................................................................... 17
7
DATA CENTRE SERVICES ..................................................................................................................... 19
7.1 7.2 7.3
DATA CENTRE SERVICES REQUIREMENTS AND CONSTRAINTS ..................................................................... 19 OTHER HE COMMUNITY DATA CENTRE INITIATIVES .................................................................................... 21 HEI PARTNERS DATA CENTRE CABINET REQUIREMENTS .............................................................................. 22
8
GOVERNANCE AND PARTNERSHIP ...................................................................................................... 23
8.1 8.2 8.3
OVERVIEW..................................................................................................................................................... 23 CONSORTIUM PARTNERSHIP MODEL ........................................................................................................... 23 JOINT VENTURE PARTNERSHIP MODEL ........................................................................................................ 23
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
7
© PTS Consulting 2010
8.4
TECHNICAL SOURCING AUTHORITY MODEL ................................................................................................. 23
9
FINANCIAL ANALYSIS ......................................................................................................................... 24
10
PROJECT CONCLUSION ....................................................................................................................... 32
10.1 OVERVIEW..................................................................................................................................................... 32 10.2 CUSTOMERS AND HEI PARTNERS.................................................................................................................. 32 10.3 DATA CENTRE SERVICE MODEL..................................................................................................................... 32 10.4 SERVICE PROVIDER PARTNERS ...................................................................................................................... 33 10.5 SHARED SERVICE PARTNERSHIP MODEL....................................................................................................... 33 10.6 FINANCIAL ANALYSIS..................................................................................................................................... 34 10.7 SUMMARY ..................................................................................................................................................... 34 11
RECOMMENDATIONS ........................................................................................................................ 35
11.1 11.2 11.3 11.4
OVERVIEW..................................................................................................................................................... 35 NATIONAL EDUCATION DATA CENTRE SERVICES.......................................................................................... 35 TRANSITION SERVICE .................................................................................................................................... 35 SERVICE MANAGEMENT ............................................................................................................................... 39
11.5 SUBJECT MATTER EXPERTISE ........................................................................................................................ 41 11.6 DATA CENTRE ACQUISITION AND SUPPLY .................................................................................................... 43 11.7 NEDCS GOVERNANCE.................................................................................................................................... 43 11.8 CONCLUSION................................................................................................................................................. 46 12
NEXT STEPS ....................................................................................................................................... 47
12.1 BUSINESS CASE DEVELOPMENT .................................................................................................................... 47 12.2 INCUBATOR PROJECT .................................................................................................................................... 48 12.3 IMPLEMENTATION ROADMAP ...................................................................................................................... 49 13
APPENDIX 1 – GLOSSARY OF TERMS ................................................................................................... 50
14
APPENDIX 2 – OBJECTIVE TIERING PROCESS........................................................................................ 53
15
APPENDIX 3 – FINANCIAL ASSUMPTIONS............................................................................................ 55
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
8
© PTS Consulting 2010
2
DOCUMENT CONTROL
2.1
DOCUMENT AUTHORITY Prepared by: Barry Lewington .................................................................................. Date: ......... 26th February 2010 Reviewed by: Martin Williams ................................................................................... Date: ............... 4th March 2010 Approved by: Martin Williams ................................................................................... Date: ............... 4th March 2010
2.2
PROJECT INFORMATION Document Number: ............................................................................................. UOSAL-02-1GEN009 Date Issued:................................................................................................................. 4th March 2010 Project Description:...................................................................................Shared Service Data Centre This Document: ................................................................................................. Final Options Analysis
2.3
CLIENT DISTRIBUTION Richard Maccabee (University of Derby) ............................................................ SSDC Steering Group Mark Hilditch (University of Salford) .................................................................. SSDC Steering Group Mary Visser (University of Leicester) .................................................................. SSDC Steering Group Paddy Walker (HEFCE) ........................................................................................ SSDC Steering Group Tim Marshall (JANET UK)..................................................................................... SSDC Steering Group
2.4
DOCUMENT HISTORY Version
Date Issued
Status
1.0
14th December 2009
Review
2.0
12th February 2010
Final
2.1
4th March 2010
Updated
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
9
Description For Steering Group Review Final Version for release Updated following client feedback
© PTS Consulting 2010
3
PURPOSE OF THIS DOCUMENT This document has been created to provide a final collective view of the Options available to the Shared Service Data Centre project team. This document summarises the conclusions of the project team in the suitability of the project progressing as scoped, and the areas of the project that would need to be further developed to benefit for the HEI community in the future. This document summarises the findings from the key stages of the project: Governance and Partnership Data Centre Services Customer Requirements Service Provider Partners Market Analysis Competitor Analysis This document also positions the present status of the project and proposes a Roadmap for the future development of the service, a proposed ‘Incubator’ project to test the model, and the benefits of continuing with such a project to the education community. The contents of this document can be used to develop a business case for the securing of funds to further develop this project for the good of the HEI community. This document can also be used as a reference point for the collection of findings from this phase of the project.
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
10
© PTS Consulting 2010
4
PROJECT DEFINITION
4.1
PROJECT BACKGROUND As part of the Higher Education Funding Council for England’s (HEFCE) Shared Service Programme, the Universities of Derby, Leicester and Salford have been developing proposals that would lead to the provision of a Shared Service Data Centre (SSDC) to meet the business needs of all 3 institutions, whilst at the same time providing a model for other Higher Education Institutions (HEIs) to follow. It was envisaged that further partner institutions would be added to the 3 primary stakeholder institutions as the proposals became more firm. It was further envisaged that the shared service data centre would, once established, provide services to other education sector institutions and possibly others outside of the education community. The University of Salford identified business continuity as its primary business driver for the development of an off‐site data centre facility. The University of Derby identified the requirement to provide a more robust primary data centre service than currently exists in order to deliver business‐critical applications to a number of geographically dispersed campuses and off‐site locations. The University of Leicester identified a need to provide a secure primary data centre facility to replace their primary data centre which was expected to be demolished as part of campus renovation work. While these practical considerations drove the inception of this project, deeper potential benefits emerged during its course. All three are presently engaged in significant change programmes within their institutions with the aim to professionalise service delivery and align IT investment more appropriately, transparently and cost-effectively to their institutional strategies. They increasingly see the opportunity to re-architect their IT infrastructure so as to exploit off-site data centre facilities as a potential catalyst of that major cultural change. They also increasingly argue that this shift could, and should, be supported by the national drive towards shared service delivery models. The project commenced in 2007 when Logica was appointed to undertake a feasibility study and to develop a business plan. Logica’s activities confirmed that a shared service data centre approach was viable and cost effective. The study also recommended that the optimum approach was to develop a new‐build facility. The project entered the next stage, to determine the options around: Customers and HEI Partners – determining what the HEI Partners business needs and desires are, and how they could be met by a shared data centre service. Discovering who else may want to make use of the service (customers), and therefore what it should offer data centre service model – discovering the services that service provider partners would be interested in offering. Determining what the future Service might consist of given the service provider partner appetite, and the needs of the HEI partners and customers. Researching the competing Services, either currently available or in development Shared partnership model – understanding how the HEI and service provider partners would prefer to work together. Discovering what may influence how they work together, and determining which partnership model options are available
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
11
© PTS Consulting 2010
4.2
PROJECT BRIEF The Universities of Derby, Leicester and Salford required the services and support of an independent specialist business and IT consultancy to develop the SSDC model to a point such that the 3 Universities and HEFCE would have a high degree of confidence regarding the viability and sustainability of the proposed model. In addition, confidence that the service provider marketplace was capable, willing, and able to respond to a formal procurement process. The outcome of these works is documented in this Options Appraisal report which makes recommendations as to the most appropriate way to proceed. As part of the recommendations this report also proposes a roadmap for the next stages to setting up a service. The latter stage is not within PTS’s scope.
4.3
PROJECT OBJECTIVES The PTS consultants have worked closely with the Universities of Derby, Leicester and Salford in: Supporting the development of the programme of works in more detail, confirming and agreeing the proposed approach and the consultant/client resource requirements Developing an appropriate Governance structure for the shared Service partnership model Leading and supporting discussions with potential Higher Education (HE) sector (and other sectors where appropriate) partners and potential customers Leading and supporting discussions with potential private sector (Private Partner) data centre service provider partners and developers Leading and supporting discussions with the Higher Education Funding Council for England (HEFCE) and JANET(UK) regarding their visions for shared service data centres, and a potential service provider role for JANET(UK) Developing the HEI Partnership, organisation and support model Developing the service delivery model, focusing on long‐term viability and sustainability Understanding the market strengths, weaknesses, threats and opportunities – including potential demand from potential customers of the service Engaging with potential service provider partners, developers and service provider partners to determine the level of interest, capabilities and willingness to enter into a tender competition, and subsequently deliver the required services
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
12
© PTS Consulting 2010
4.4
PROJECT APPROACH Figure 1 graphically shows the approach that PTS undertook to complete the Options phase of the SSDC initiative.
Figure 1 – PTS Project Approach Table 1 details the key questions that this project attempts to answer, and within which stage of the project. Question
Stage
PRIMARY What were the business needs and desires of the HEI partners that could have been met by a shared data centre Service?
Stage 2: HEI Partners and Customers
SECONDARY Who else may want to make use of the service and what it could offer? PRIMARY Which services could be provided – i.e. platform as a service, infrastructure as a service, co-location?
Stage 3: Data Centre Service Model
SECONDARY What competing services are currently available? PRIMARY How could the service provider partners and HEI partners work together?
Stage 4: Shared Service Partnership Model
SECONDARY What are the appropriate governance model options (inclusive of finance and tax)? PRIMARY Which combination of options (partnership model, data centre service, and HEI partners/customers) work, and which preferred options would be taken forward?
Stage 5: Options Appraisal
SECONDARY What would the costs, timescales, risks and benefits of the identified options be? Table 1 - Project Brief SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
13
© PTS Consulting 2010
5
CUSTOMER REQUIREMENTS In assessing and capturing the customer requirements, 3 customer groups were explored by PTS: the existing HEI partners (Universities of Derby, Leicester and Salford) potential customers identified and proposed by the 3 HEI partners results of a HE and FE community survey conducted by PTS in conjunction with JANET UK
5.1
HEI PARTNER REQUIREMENTS The 3 partner’s requirements were identified as requiring core data centre services (colocation, disaster recover, and infrastructure services). It was also determined that their requirements would occur at different times over a period of 4 years: The University of Derby would require services immediately to provide primary data centre services The University of Leicester would require services within 2 to 3 years when internal infrastructure work had been completed The University of Salford would require services in 3 to 4 years as a result of campus redevelopment work requiring the demolition of the existing primary data centre facility The total requirements from all 3 institutions were presently stated (if services were migrated as is, without any further virtualisation work being undertaken) as: Total cabinet requirements – 245 cabinets Total server requirements – 1 490 server instances Consolidating these requirements, and making the following assumptions: all server instances (presently Windows platforms) could and would be migrated to a virtualised state server infrastructure would be of high density technology server chassis would serve a maximum of 60 server instances a data centre cabinet would store 3 server chassis It was estimated that the 3 institutions could expect to be able to migrate their primary IT services to a data centre with approximately 452 cabinets (servers, storage and network).
2 It should be noted that these figures are based upon theoretical assumptions and estimates, providing an indication only. A full service assessment should be carried to fully assess the suitability of services to be migrated onto a virtualised environment SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
14
© PTS Consulting 2010
5.2
POTENTIAL CUSTOMERS The project team interviewed a number of potential customers identified by the 3 HEI partners. Each potential customer had differing requirements as follows: University of Southampton – shortage of existing space requiring immediate data centre capacity Loughborough University – shortage of suitable data centre space, data centre end-of-life, requiring immediate data centre capacity Anglia Ruskin University – looking for a secondary data centre to cover its Peterborough campus requirements University of West of England – will require a data centre facility when the IT department is moved onto a new campus (have a potential site for a new data centre) Peel Media – has a new data centre facility and was looking for someone to manage this facility The conclusion of this activity was that there is an immediate requirement for data centre services with institutions in end-of-life facilities, or lacking space or power to implement new services.
5.3
REQUIREMENTS IDENTIFIED FROM SURVEY OF HEI AND FE COMMUNITIES PTS, in conjunction with JANET (UK), conducted a survey of 163 HEIs through invitation to complete an online survey response. Sixty-six (40.5%) institutions responded to the survey. The aim of the survey was to understand the: Institutions plans for data centre services and which options they would consider for their future data centre Financial breakdown of data centre services today and any predicted changes over the coming years Power utilisation breakdown of the existing data centre Level of risk acceptance within the institutions for a shared service data centre Future delivery plans for common education applications i.e. email, VLE etc. The responses showed that: Of the 66 respondents 61% would look to participate in a shared service, and of those 45% would look to take part within the next 2 years, with 50% taking part in 3 - 5 years On exploring the service needs of the institutions, the respondents stated virtualisation and disaster recovery were immediate requirements When respondents were asked about their IT budget only 29% were either able or willing to respond with a high level estimate of their annual IT budget Over 70% of respondents do not run any form of capacity forecasting for their IT services When respondents were asked for their opinions on the benefits a shared service would provide, in general the majority felt that the benefits would be better than their current service When respondents were asked about their power consumption, few were able to respond accurately due to the usage of power either not being monitored, or not broken out from general campus power usage
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
15
© PTS Consulting 2010
When respondent were asked about the various risks that a shared service may bring to their existing service provision, the majority of respondents felt that the risks were acceptable Respondents were asked about their existing and future thoughts with regard to the delivery of key institution applications. The majority responded stating that in the future key applications would be delivered internally, except student email which would likely be outsourced
5.4
TRANSITION REQUIREMENTS The project team identified the transition challenge as a key risk and area of uncertainty and doubt for institutions considering migrating services to a new data centre facility. Transition of services covers a comprehensive process of identifying the components of a service, the future requirements of the service, and identifying and documenting a process for the successful transition of a service into a new data centre facility. With such a complex requirement PTS uses an established 9 stage process: 1. Pre-migration consultancy and scoping 2. Application and service mapping benchmarking 3. Customer service requirements 4. Service catalogue generation 5. New service requirements generation 6. Service procurement 7. Service migration planning 8. Service migration activity 9. Retirement and decommission of old services The transition skills requirements would provide the HEIs with the support and access to specialist skills to develop the transition programmes. Transition service requirements were identified as part of several ‘consultancy’ services that would assist the HEIs in identifying, communicating and improving the delivery of IT data centre services to their institutions.
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
16
© PTS Consulting 2010
6
SERVICE PROVIDER PARTNERS Service provider partners were viewed by the project team as important to the future success of the project. Service provider partners were identified as being able to bring a number of important benefits that the existing HEI partners felt they were unable to provide themselves, in particular: Finance – a service provider may be invited to join the SSDC partners as a Private Partner in a Joint Venture to provide a service to a wider community of institutions, with the Private Partner investing financially to help set up the Joint Venture business Technical Skills – the service provider partners would be key in bringing up to date technical skill-sets to the service which were seen as generally lacking in the institutions today Infrastructure – all 3 HEI partners stated that they would not be migrating infrastructure to the shared service and so would look at key strategic service provider partners to provision core infrastructure (network, servers, storage) for the delivery of services Management – the partners would look at the service provider community to provide the overall governance and management of the service to the HEI partners The Project team met with a list of identified service provider partners (see Table 2) to understand their appetite to participate in the project, and to explore interest in investing in the project within a Joint Venture partnership. The project team approached 17 service provider partners covering all elements of data centre service including: Overall day-to-day management of the service Delivery of a physical data centre or data centre space Delivery of a network platform Delivery of a server platform Delivery of server system operating software (traditional and virtualised) Delivery of storage platforms All service provider partners were informed that the meetings were an exploratory meeting and that the next stages would likely be via a formal procurement route. In nearly all cases the service provider partners were enthusiastic about the project potential and would welcome the opportunity to participate in the project. Most of the service provider partners were willing to explore ways of supplying their services within a flexible reward system. Few service provider partners showed an interest in a Joint Venture Partnership, preferring a utility3 type service model for the supply and reward of their services. It was clear from discussions that the service provider community would need to gain executive management sponsorship for the project if they were to invest time and resources in the supply of infrastructure. This would take more time with some service providers, and would need a more formal engagement to continue discussions i.e. via a Request for Information (RFI) or Invitation to Tender (ITT). Table 2 below details the Service Providers that were met and the services that they provide.
3 Utility Services are where the service provider provides their services or infrastructure platform without any initial reward and the customer only pays the service provider for the services they use SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
17
© PTS Consulting 2010
Service Provider Partner
Services Provided
Cisco
The provision of networking and the new UCS (Unified Computing System)
Cisilion
A channel partner for the Modulco/Bladeroom flexible data centre unit
COLT
The provision of pre-built data centre space
Dell
The provision of server and storage infrastructure
Eduserve
The provision of pre-built data centre space
EMC
The provision of storage and VMware software services, also in partnership with Cisco under the VCE partnership
Fujitsu
The provision of end to end services including the infrastructure management also able to deliver: pre-provisioned data centre space, servers, storage
Getronics
The provision of the infrastructure management and able to provide pre-provisioned data centre space
Hitachi Data Systems
The provision of storage services
HP
The provision of end to end services including the infrastructure management also able to deliver: pre-provisioned data centre space, servers, storage and network
IBM
The provision of end to end services including the infrastructure management also able to deliver: pre-provisioned data centre space, servers, storage and network
Kcom
The provision of pre-built data centre space
Microsoft
The provision of software services
Mitie
The provision of custom built data centres plus the ongoing maintenance of the Mechanical and Electrical element of these facilities
Moduleco/BladeRoom
The provision of portable flexible data centre units, plus the maintenance of the Mechanical and Electrical elements of these units
Sun Microsystems
The provision of server and storage infrastructure
Sungard
The provision of pre-built data centre space Table 2 - Service Providers
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
18
© PTS Consulting 2010
7
DATA CENTRE SERVICES
7.1
DATA CENTRE SERVICES REQUIREMENTS AND CONSTRAINTS In determining the services a data centre should provide, the HEI partners, as customers, were interviewed and their requirements captured. Other potential customers were also interviewed and their requirements captured, and the wider HEI community was surveyed on its general needs as a community. The service provider partners were also interviewed to assess the types of services they would be able to provide in addressing the HE community’s needs. In taking the captured needs of the HEI community, and the scope of services the service provider partners were able to provide, the project team was able to develop a data centre services model. The following key issues were identified as requiring a solution from the delivery of services within the HEI data centre: Lack of power Lack of cooling Lack of space Construction work within the campus necessitating the demolishment of an existing data centre / estates department plans Lack of technical skills Move to virtualised platforms Lack of disk storage Infrastructure equipment at end of life No primary/secondary data centre (no disaster recovery) Whereas these primary drivers brought each institution to the table, several additional requirements emerged during the project with equal importance: The need to improve IT service delivery cost effectiveness and be able to demonstrate best value in a way that the business understands The need to drive step change in service delivery culture The need to design services to meet availability, capacity and business continuity requirements that are transparently negotiated with the business The need to realign the workforce away from ‘keeping the lights on’, to roles that add value directly to the teaching and research activities of the institution The need to reduce carbon footprint and provide carbon tax readiness and efficiency The need to improve IT governance within institutions In addressing these key requirements the data centre was seen to offer a scalable range of services to support the drive for change in their institutions. Initially providing basic services (colocation space), it is expected that it would grow into more advanced services (virtualised server and storage). It was agreed by the Project Steering Group that the scope of services delivered should initially stop at the delivery of infrastructure services. The delivery of shared applications is under consideration in other initiatives. The Shared Service Architecture (Figure 2) was created to show the scalable levels of service that the shared service data centre could deliver.
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
19
© PTS Consulting 2010
At a basic level was the need to create a data centre space – this could be delivered through a number of different options: Construction of a new facility either on or off campus Construction and procurement of a modular data centre facility Refurbishment of an existing data centre facility Procurement of existing data centre space from a private service provider Although the project team initially explored a single solution for the supply of a single data centre, it was concluded that the source and location of the data centre was not critical and that a final solution may be made up of space from a number of sources. The only critical item was that each data centre would require high capacity resilient network connections to the JANET network. At this stage of the project the team did not determine the minimum requirements for the sourcing of data centre space. The overall data centre service should address the Green Agenda and where appropriate the data centre should: 1. Source the power from renewable energy sources 2. Provide an efficient data centre with a Power Utilisation Efficiency (PUE) <1.3 3. Ensure that the infrastructure within the data centre was efficient and, wherever possible, powered down when not in use 4. Consider, though not mandatory, the recycling or reuse of waste heat from the data centre (i.e. heating sports facilities, greenhouses, public facilities or local housing)
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
20
© PTS Consulting 2010
Figure 2 - Shared Service Data Centre - Architecture Framework In delivering the infrastructure, the data centre would provide a range of platforms, encouraging customers to subscribe to services further up the infrastructure scale. In using such a model, replicated in different areas of the country, services could be backed up across data centres with critical services either load-balanced across multiple data centres, or disaster recovery services provided by a partner data centre. This could enable services to be transparently switched across to a different data centre in the event of an unexpected or planned data centre outage. Existing customers would have to undertake considerable work to position their current services such that they are ready to be migrated into a shared data centre service. Over time there could be several shared service data centre partnerships developed, each delivering data centre services using the shared service data centre model.
7.2
OTHER HE COMMUNITY DATA CENTRE INITIATIVES The Shared Service Data Centre project is one of four initiatives presently exploring data centre services for the HEI, the four being: Shared Service Data Centre (SSDC) – Universities of Derby, Leicester and Salford Shared Virtual Data Centre (SVDC) – Yorkshire, Humberside Metropolitan Area Network (YHMAN) Shared High End Data Centre (SHED) – Russell Universities Group IT Directors Forum UK Wide Education Sector Data Centre Strategy – JANET UK
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
21
© PTS Consulting 2010
7.3
HEI PARTNERS DATA CENTRE CABINET REQUIREMENTS In assessing the immediate needs for the 3 partners of the SSDC project, a high level view was undertaken to assess the future data centre requirements. A number of assumptions were made in developing the model. The present requirements from the 3 partners for primary data centre services were assumed as: Presently use 245 cabinets 1490 server instances All servers could be virtualised (in practice this may not be the case) Standard storage requirements Location was not an issue The following assumptions have been used in calculations based upon best practice knowledge: A blade server chassis could be virtualised to run up to 100 server instances A cabinet can contain up to 3 blade server chassis A virtualised storage array for the partners would use less than 3 cabinets Taking into account the above requirements and assumptions it is clear that an approximate total of less than 10 cabinets would be required for the 3 partners (taking into consideration networking equipment and servers that could not be virtualised also moving to the new data centre). However, this is a theoretical minimum, therefore it was agreed a more pragmatic figure would be 45 based upon understanding (untested) of what could be virtualised. If the theoretical minimum, or near to it, was possible the limited number of populated cabinets (10) would mean that the existing SSDC partners would not need to develop a data centre facility dedicated solely to their own use.
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
22
© PTS Consulting 2010
8
GOVERNANCE AND PARTNERSHIP
8.1
OVERVIEW The project team explored Governance and Partnership Models suitable for the provision of data centre services to the 3 partners and invited customers. The project team concluded that 3 options were suitable for the effective governance of a delivery model. These are: 1. A Consortium Partnership Model 2. Joint Venture Partnership Model 3. Technical Sourcing Authority Model Each option is explored in further detail below.
8.2
CONSORTIUM PARTNERSHIP MODEL The Consortium Partnership Model would consist of a partnership made up of: The University of Derby The University of Leicester The University of Salford JANET UK / HEFCE The Partnership would develop the strategy for the service, procuring services from specialist service provider partners. The Consortium would require an experienced service provider partner to provide the day-to-day management of services. This would be achieved through a competitive procurement selection process. The individual services within the data centre facility would be procured from individual service provider partners under a service contract basis.
8.3
JOINT VENTURE PARTNERSHIP MODEL The Joint Venture Partnership Model consists of the 3 partners developing a partnership under a legal entity model (Limited Company), selecting a suitable service provider partner through a standard procurement framework, and developing a joint venture between the partners and the service provider organisation. The key benefit of the joint venture model is that this allows both parties (the HEI partners and the service provider organisation) to invest finance into the joint venture under a contractual framework, with both parties recognising the risks and rewards of the venture.
8.4
TECHNICAL SOURCING AUTHORITY MODEL The Technical Sourcing Authority Model provides a model where one organisation acts a central procurement sourcing partner. By collectively pooling the HEI partner’s requirements, the HEI partners can gain improved pricing through economies of scale. This model assumes that services would only be purchased for organisations within the partnership. There is no legal entity created for this model. This model would also assume that all the partners are at similar stages in the development of their IT data centre services so that services could be purchased collectively.
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
23
© PTS Consulting 2010
9
FINANCIAL ANALYSIS During the first stage of the SSDC Project, a Financial Analysis was undertaken by Logica to demonstrate the potential savings that could be made from a shared service. In developing this next stage of the SSDC project, the financial analysis has been updated. Assumptions previously used have been updated, and additional assumptions added to develop an accurate and current analysis of the opportunities from the implementation of such a project. A list of the detailed assumptions can be found in Appendix 3 – Financial Assumptions. The financial analysis does not account for the following running costs: Day-to-day management and support of the IT infrastructure (tools and people) Day-to-day management and support of the Data centre facility (tools and people) Procurement of IT infrastructure The following provides an overview of the assumptions used in calculating the various Service Option available to the partners as detailed in the 10 year Financial Analysis (see Table 3) Option A - Bring Existing Facilities to a Baseline Level and Meet Expansion Requirements Accounts for: The cost of bringing the existing facilities up to date and developing facilities to meet expected growth That growth is based on the existing IT architecture That power costs will rise at 10% higher than inflation That the facility will be maintained at a cost of 10% of initial capital outlay That the cost of capital borrowing is in line with institution policies That IT infrastructure costs are not considered for this exercise That staffing costs are not considered for this exercise That the following institution assumptions made Leicester will start at 120 cabinets and rise to 165 Derby will start at 65 cabinets and rise to 92 Salford will start at 60 cabinets and rise to 105 That the average power draw will be 3kw/hr per cabinet That the power utilisation efficiency of the facility would be 1.6 That IT services will remain in the facility while facility upgrades take place That existing JANET connections would be maintained
Option A(a) - Bring Existing Facilities to a Baseline Level and Meet Expansion Requirements with aggressive Virtualisation Accounts for : The cost of bringing the existing facilities up to date and developing facilities to meet expected growth That growth is based on an aggressive Virtualised IT architecture That power costs will rise at 10% higher than inflation SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
24
© PTS Consulting 2010
That the facility will be maintained at a cost of 10% of initial capital outlay That the cost of capital borrowing is in line with institution policies That IT infrastructure costs are not considered for this exercise That staffing costs are not considered for this exercise That the following institution assumptions made Leicester will start at 50 cabinets and rise to 77 Derby will start at 15 cabinets and rise to 24 Salford will start at 15 cabinets and rise to 24 That the average power draw for Derby and Salford will be 9kw/hr per cabinet (Leicester will have a power draw of 6Kw due to the mix of equipment used) That the power utilisation efficiency of the facility would be 1.6 That IT services will remain in the facility while facility upgrades take place That existing JANET connections would be maintained
Option B(1) - Pursue Individual Solutions - New Build - existing IT Accounts for: The cost of constructing a data centre facility to Tier II standard for EACH institution That growth is based on the existing IT architecture That power costs will rise at 10% higher than inflation That the facility will be maintained at a cost of 10% of initial capital outlay That the cost of capital borrowing is in line with institution policies That IT infrastructure costs are not considered for this exercise That staffing costs are not considered for this exercise That the following institution assumptions made Leicester will start at 120 cabinets and rise to 165 Derby will start at 65 cabinets and rise to 92 Salford will start at 60 cabinets and rise to 105 That the average power draw will be 3kw/hr per cabinet That the power utilisation efficiency of the facility would be 1.3 That a there is a cost for each institution to migrate services from their existing data centre to the new That existing JANET connections would be maintained
Option B(1)a - Pursue Individual Solutions - New Build - with aggressive Virtualisation Accounts for: The cost of constructing a data centre facility to Tier II standard for EACH institution on existing institution land That growth is based on an aggressive Virtualised IT architecture That power costs will rise at 10% higher than inflation That the facility will be maintained at a cost of 10% of initial capital outlay That the cost of capital borrowing is in line with institution policies SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
25
© PTS Consulting 2010
That IT infrastructure costs are not considered for this exercise That staffing costs are not considered for this exercise That the following institution assumptions made Leicester will start at 50 cabinets and rise to 77 Derby will start at 15 cabinets and rise to 24 Salford will start at 15 cabinets and rise to 24 That the average power draw for Derby and Salford will be 9kw/hr per cabinet (Leicester will have a power draw of 6Kw due to the mix of equipment used) That the power utilisation efficiency of the facility would be 1.3 That a there is a cost for each institution to migrate services from their existing data centre to the new That existing JANET connections would be maintained in the institutions
Option B(2) - Pursue Individual Solutions - Co-location - existing IT Accounts for: The cost for EACH institution undertaking a colocation contract with a private service provider That growth is based on the existing IT architecture That power costs will rise at 10% higher than inflation That annual costs are based on the rental of cabinet space plus power utilisation That the cost of capital borrowing is in line with institution policies That IT infrastructure costs are not considered for this exercise That staffing costs are not considered for this exercise That the following institution assumptions made Leicester will start at 120 cabinets and rise to 165 Derby will start at 65 cabinets and rise to 92 Salford will start at 60 cabinets and rise to 105 That the average power draw will be 3kw/hr per cabinet That the power utilisation efficiency of the facility would be 1.6 That a there is a cost for each institution to migrate services from their existing data centre to the new That a JANET connections would be installed in EACH colocation facility to set up a connection to the institution
Option B(2)a - Pursue Individual Solutions - Co-location - with aggressive Virtualisation Accounts for: The cost for EACH institution undertaking a colocation contract with a private service provider That growth is based on an aggressive Virtualised IT architecture That power costs will rise at 10% higher than inflation That annual costs are based on the rental of cabinet space plus power utilisation That the cost of capital borrowing is in line with institution policies SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
26
© PTS Consulting 2010
That IT infrastructure costs are not considered for this exercise That staffing costs are not considered for this exercise That the following institution assumptions made Leicester will start at 50 cabinets and rise to 77 Derby will start at 15 cabinets and rise to 24 Salford will start at 15 cabinets and rise to 24 That the average power draw for Derby and Salford will be 9kw/hr per cabinet (Leicester will have a power draw of 6Kw due to the mix of equipment used) That the power utilisation efficiency of the facility would be 1.6 That a there is a cost for each institution to migrate services from their existing data centre to the new That a JANET connections would be installed in EACH colocation facility to set up a connection to the institution
Option C(1) - Shared Service - 3 Institutions - New Build - existing IT Accounts for: The cost of constructing a shared data centre facility to Tier II standard for the partner institutions on existing institution land That growth is based on the existing IT architecture That power costs will rise at 10% higher than inflation That the facility will be maintained at a cost of 10% of initial capital outlay That the cost of capital borrowing is in line with institution policies That IT infrastructure costs are not considered for this exercise That staffing costs are not considered for this exercise That the following institution assumptions made Leicester will start at 120 cabinets and rise to 165 Derby will start at 65 cabinets and rise to 92 Salford will start at 60 cabinets and rise to 105 That the average power draw will be 3kw/hr per cabinet That the power utilisation efficiency of the facility would be 1.3 That a there is a cost for each institution to migrate services from their existing data centre to the new That a new resilient JANET connection would be implemented
Option C(1)a - Shared Service - 3 Institutions - New Build - with aggressive Virtualisation Accounts for: The cost of constructing a shared data centre facility to Tier II standard for the partner institutions on existing institution land That growth is based on an aggressive Virtualised IT architecture That power costs will rise at 10% higher than inflation That the facility will be maintained at a cost of 10% of initial capital outlay That the cost of capital borrowing is in line with institution policies SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
27
© PTS Consulting 2010
That IT infrastructure costs are not considered for this exercise That staffing costs are not considered for this exercise That the following institution assumptions made Leicester will start at 50 cabinets and rise to 77 Derby will start at 15 cabinets and rise to 24 Salford will start at 15 cabinets and rise to 24 That the average power draw for Derby and Salford will be 9kw/hr per cabinet (Leicester will have a power draw of 6Kw due to the mix of equipment used) That the power utilisation efficiency of the facility would be 1.3 That a there is a cost for each institution to migrate services from their existing data centre to the new That a new resilient JANET connection would be implemented
Option C(2) - Shared Service - 3 Institutions - Co-location - existing IT Accounts for: The cost for the partner institutions to undertake a colocation contract with a private service provider That growth is based on the existing IT architecture That power costs will rise at 10% higher than inflation That annual costs are based on the rental of cabinet space plus power utilisation That the cost of capital borrowing is in line with institution policies That IT infrastructure costs are not considered for this exercise That staffing costs are not considered for this exercise That the following institution assumptions made Leicester will start at 120 cabinets and rise to 165 Derby will start at 65 cabinets and rise to 92 Salford will start at 60 cabinets and rise to 105 That the average power draw will be 3kw/hr per cabinet That the power utilisation efficiency of the facility would be 1.6 That a there is a cost for each institution to migrate services from their existing data centre to the new That a JANET connections would be installed in the colocation facility to set up a connection to the institution partners
Option C(2)a - Shared Service - 3 Institutions - Co-location - with aggressive Virtualisation Accounts for: The cost for partner institutions to undertake a colocation contract with a private service provider That growth is based on an aggressive Virtualised IT architecture That power costs will rise at 10% higher than inflation That annual costs are based on the rental of cabinet space plus power utilisation That the cost of capital borrowing is in line with institution policies SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
28
© PTS Consulting 2010
That IT infrastructure costs are not considered for this exercise That staffing costs are not considered for this exercise That the following institution assumptions made Leicester will start at 50 cabinets and rise to 77 Derby will start at 15 cabinets and rise to 24 Salford will start at 15 cabinets and rise to 24 That the average power draw for Derby and Salford will be 9kw/hr per cabinet (Leicester will have a power draw of 6Kw due to the mix of equipment used) That the power utilisation efficiency of the facility would be 1.6 That a there is a cost for each institution to migrate services from their existing data centre to the new That a JANET connections would be installed in the colocation facility to set up a connection to the partner institutions In reviewing the financial analysis summary (Error! Reference source not found.) the key areas that have driven changes from the original Logica analysis are the: effects of increasing costs of power during the course of the study effects of efficiency within data centres in the reduction in power utilisation potential opportunities in reducing costs through the implementation of efficient IT technologies such as Virtualisation (detailed evaluation of virtualisation technologies are out of scope for this project, however, the potential impact upon data centre requirements has been accounted for in the financial model) – these can be demonstrated by the savings made between updating the existing facility with the existing IT infrastructure and through using aggressive virtualised infrastructure (Savings of £4.25mill over the 10 year period) cost impact that transition to a new data centre will incur on the project variable costs of data centre facilities, such as variances in data centre location and facility quality (i.e. tier level) requirement for a greater number of institutions to create the economies of scale savings increasing costs of the colocation model due to customer and legislative demands, and power cost increases real advantages of a ‘Green Data Centre’ not just the reduction in power usage, but also cost savings in areas such as carbon tax due to the costs of construction and migration of services, the cost of developing new data centres for each of the institutions weighed against upgrading existing facilities demonstrate a higher cost of running services (Additional £44m over the 10 year period) – this is due to the smaller size of the facilities and the inefficient use of IT infrastructure which outweighs any advantages the data centre facility can bring to the service through the use of higher density IT infrastructure the cost of construction of the new facilities are reduced and improved efficiencies in running the facility can be seen with a small 10 year saving of £400k if the institutions explored the use of colocation services for each of the estates, the inhibited cost of setting up new JANET networks and the cost of installing and annual rental of cabinets. It should be noted that institutions should also expect to pay a higher premium for power in developing a colocation service using a virtualised IT infrastructure, the institutions should expect to pay a premium for their cabinets due to the above average requirement for cooling of platforms. This can be outweighed by the reduced number of cabinets overall
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
29
© PTS Consulting 2010
and this is demonstrated by the large saving against existing infrastructure architecture within a colocation. This is though more expensive that updating the existing facility in considering a shared data centre facility based on existing IT infrastructure, as considered in 2 the cost of constructing a large facility is outweighed by the cost of maintaining inefficient IT. There is also an additional cost of implementing a larger JANET network connection to provide links back to the partner institutions in considering a shared data centre with the implementation of efficient IT, greater savings can be achieved with a 10 year saving of £9.8 mill in considering the three partner institutions sharing a colocation service it was concluded that there were few savings to be found over and above the institutions exploring their own colocation services. The only savings found would be in the sharing of a JANET network connection out of the shared colocation facility The analysis provided takes into account the data centre facility requirements, and migration costs. However, it does not take into account the costs of the next layer of infrastructure, for example network, server and storage architectures, and therefore the associated cost savings in sharing these components across institutions under a shared service. Inclusion of these elements would put further weight behind the use of a shared data centre service, however, it is not possible at this stage to determine the level of sharing that would be achievable. The high level review within this project explored the need for the effective design, implementation and management of a shared infrastructure platform; the assumption is that this will be the responsibility of a selected service provider.
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
30
© PTS Consulting 2010
Adjusted NPV
Original Net Present Value
Risk / Bias
Annual Cost over 10 years
Average Annual Cost Per Institution
Increase in Av. Annual Cost Per Institution
NPV relative to Option A
Option A Bring Existing Facilities to a Baseline Level and Meet Expansion Requirements
£50,855,142
£49,295,142
£1,560,000
£5,085,514
£1,695,171
£0
£0
Option A (a) Bring Existing Facilities to a Baseline Level and Meet Expansion Requirements - with aggressive Virtualisation
£46,599,767
£45,039,767
£1,560,000
£4,659,977
£1,553,326
-£141,846
-£4,255,375
Option B (1) Pursue Individual Solutions - New Build - existing IT
£94,070,253
£91,732,253
£2,338,000
£9,407,025
£3,135,675
£1,440,504
£42,437,111
Option B (1a) Pursue Individual Solutions - New Build - with aggressive Virtualisation
£50,924,520
£48,586,520
£2,338,000
£5,092,452
£1,697,484
£2,313
-£708,622
Option B (2) Pursue Individual Solutions - Co-location - existing IT
£108,332,261
£105,229,261
£3,103,000
£10,833,226
£3,611,075
£1,915,904
£55,934,119
Option B (2a) Pursue Individual Solutions - Co-location - with aggressive Virtualisation
£56,253,863
£53,150,863
£3,103,000
£5,625,386
£1,875,129
£179,957
£3,855,721
Option C (1) Shared Service - 3 Institutions - New Build - existing IT
£57,044,550
£55,521,550
£1,523,000
£5,704,455
£1,901,485
£206,314
£6,226,408
Option C (1a) Shared Service - 3 Institutions - New Build - with aggressive Virtualisation
£40,977,036
£39,454,036
£1,523,000
£4,097,704
£1,365,901
-£329,270
-£9,841,105
Option C (2) Shared Service - 3 Institutions - Co-location - existing IT
£97,156,016
£93,303,016
£3,853,000
£9,715,602
£3,238,534
£1,543,362
£44,007,874
Option C (2a) Shared Service - 3 Institutions - Co-location - with aggressive Virtualisation
£54,172,126
£50,319,126
£3,853,000
£5,417,213
£1,805,738
£110,566
£1,023,984
10 Year Financial Analysis – Based on Disposal of Current Data Centre Facilities and use of JANET Infrastructure
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
31
© PTS Consulting 2010
10
PROJECT CONCLUSION
10.1
OVERVIEW In concluding the 3 key phases of this project (Customers and HEI Partners, Data Centre Service Model and Shared Service Partnership Model) the project team determined that the interrelationship between the identified phases were inseparable, with output from one phase driving the development and conclusion of the others.
10.2
CUSTOMERS AND HEI PARTNERS
In concluding the Customer Requirements phase it is clear that institutions have immediate requirements for data centre services. Some institutions have immediate requirements and would look favourably upon a market sector focused organisation centred upon the delivery of data centre services. However, present requirements for data centre services tend to be focused in the low risk areas such as disaster recovery services. Where institutions have already completed work in the virtualisation of services, these could easily port to a common shared infrastructure platform. The results from the national survey undertaken indicate that there is a general scepticism by institutions for the migration of services to third parties. PTS believes that this will improve as the maturity of services delivered by service providers within the education sector increases. Witnessing this trait within other market sectors, PTS has seen the cultural acceptance of such services evolve and expects any scepticism to diminish over time. It has been identified that the key stumbling block to transition to a shared service is the institutions themselves, their ability to encompass change, and their service management maturity. Winning the case for change is also a significant challenge for IT leaders within the institutions. The transition skills required from the shared service must include support and access to specialist skills in order to develop the institution specific business cases and transition programmes. The conclusion, therefore, is that there is a business requirement for a shared data centre service within the HEI sector, and clear long-term cost benefits. A range of data centre specific services should be provided that are competitive, reducing the overall cost of services to institutions and benefitting from improvement to the level of service quality. The service requirements must not be limited to the provision of infrastructure components and support, but should also extend to essential transition support and business case development.
10.3
DATA CENTRE SERVICE MODEL To identify the requirements for the data centre services the project team required a customer base to identify and capture their initial service needs. Through a combination of meetings, seminars and a national online survey, the project team were able to collect invaluable data to model a service that would meet the identified customer demands. It was identified from the customer requirements analysis that the primary benefit from the implementation of a Shared
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
32
© PTS Consulting 2010
Service Data Centre to a wider community would be the increased benefits of quality, availability and flexibility of service, plus, the community identified access to current skill sets were key to improving the delivery of the institutions services. In concluding the data centre services phase, it was agreed that the SSDC Partnership would not benefit from developing a data centre of its own, and that its requirements would best be delivered via a brokerage service from a central group pooling services nationally. This national group would acquire/procure services utilising an appropriate procurement framework. The national group would then manage and sell these services back into the sector in a similar fashion that JANET UK does with network services today. This National Education Data Centre Services (hereafter referred to as NEDCS – working title) would retain the skill sets required to effectively procure data centre services and drive cost efficiencies and service improvement into the HE sector, whilst also defining the minimum service requirements from data centre service providers.
10.4
SERVICE PROVIDER PARTNERS In concluding the service provider partner phase of the SSDC project, it was clear that there was a very supportive service provider partner community eager to participate in the project. Almost all service provider partners were willing to participate within a utility model, and so a service could be built from the service provider partners approached, with little up front finance needed from the HEI partners to start up services. It would, however, be advisable to first explore a more detailed pricing model from the service provider community which could be obtained through the issue of a detailed Request for Information (RFI) process. The project team was unable to identify suitable third party partners without identifying the services that would be required to be delivered from within the data centres first. It was agreed by the project team though that all services delivered from within the data centre would be delivered by selected service providers and that no existing institution staff would be utilised within the data centre. It was also agreed that all infrastructure required within the data centre would be provided by the service provider community. Initially services would be limited to infrastructure services only, with the scope expanded in the future to include application and software collaboration. The infrastructure platforms could be developed in conjunction with the partners further to provide scalable virtual server instances and tiered storage services. The project team identified the requirement to deliver a best of breed infrastructure platform, which could be delivered with service providers rewarded through a utility service model, where the service provider delivers and manages the infrastructure platforms and institutions subscribe to the service on a pay as you use basis. It was agreed that greater benefits could also be gained through the introduction of private sector service provider organisations contracted to introduce and deliver best practices in the operation and management of the data centres which could yield improved quality, greater efficiencies and improved availability of services.
10.5
SHARED SERVICE PARTNERSHIP MODEL In identifying the governance and partnership model, many options were explored, but concluded that a flexible solution of the partners working together and developing a relationship with one or more private sector partners to deliver the services required would enable a mutually beneficial solution for all parties. However, it was agreed by the project team that this was no longer required as a separate partnership as future services should be sourced from a national service (NEDCS) focused upon
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
33
© PTS Consulting 2010
the supply of data centre services to the education community. The existing HEI partner’s requirements would be assumed to be the same as any other HEI customer.
10.6
FINANCIAL ANALYSIS In conclusion of the financial analysis it is evident that cost savings would be made if the HEIs made use of a shared data centre. However, these cost savings are based upon aggressive virtualisation and use of skills and expertise provided by Service Providers. The study demonstrated significant savings over a ten year period. However, looking at the financial analysis alone is misleading as the results of the project also showed that the scope and timing of the 3 HEI partners’ requirement were unlikely to bring about any significant vendor leverage.
10.7
SUMMARY In concluding the Final Options stage, it was identified that if the 3 partners were to develop the solution without an additional customer base, and that the 3 partner’s institutions had implemented and completed their virtualisation projects, the IT Infrastructure volume requirements were deemed to be insufficient to attract the service provider community to participate in the project. It was also identified that the provision of a data centre and infrastructure alone was insufficient to take the institutions forward in meeting their business requirements. The project team identified the need for a national education focused organisation to provide the community with access to a breadth of data centre skills and services. A national organisation could provide access to the education community to data centre skills and services. The national organisation would procure the services required to enable organisations to migrate their data centre infrastructure from their existing facilities to modern efficient facilities managed by selected service providers under agreed levels of service at competitive market rates. Through the procurement of services using economies of scale, the education community would benefit, through reduction in operating costs, improvement in availability and continuity of services, access to industry and sector specific skill-sets, and reduction in delivery timescales. In closing, there does though still remain a level of scepticism within the community regarding the potential value and success that a Data Centre Shared Service could bring to the community. The project team strongly recommend the implementation of an ‘Incubator Project’ focused on delivering a clearly defined and scaled down service to a limited number of early adopter clients. The key target of the project could be to prove the financial numbers within the business plan and to demonstrate the scope and reliability of services that could be provided by such a service and alleviate some of the scepticism while generating a momentum that would develop the long term plans for a successful community service.
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
34
© PTS Consulting 2010
11
RECOMMENDATIONS
11.1
OVERVIEW The service requirements of the institutions expand further than just the provision of a data centre and include the requirement for access to technical and transition expertise services. It was also determined that the three HEI partners did not have requirements of a scale that would bring about the savings desired. In developing the solution for the three HEI partners it was determined that the requirements would best be met by a national data centre service. In considering this scope, the SSDC project team concluded a requirement for a centralised education sector focused organisation to provide a range of data centre services was in the best interest of the Higher Education sector. The SSDC project team has used a working title of National Education Data Centre Service (NEDCS) to provide reference and clarity throughout the remainder of this document. In identifying the role of this organisation it was clear that the key objective is to effectively source, and provide access to, the technical and sector experience and expertise required. The NEDCS would be efficiently resourced with the target to provide a route to industry specialists at competitive costs. The NEDCS would procure services through standard procurement frameworks passing on economy of scale savings to its customers – education institutions.
11.2
NATIONAL EDUCATION DATA CENTRE SERVICES In assessing the services that should be provided to the community, the SSDC partners identified a number of key services that would need to be sourced to enable them to provide a data centre service to their institution. The service structure would be similar that used by JANET UK to deliver the UK Education Network. The SSDC partners identified the following services as a minimum requirement from the NEDCS: Change Consultancy Services Service Management Subject Expertise Data Centre Acquisition/Supply
11.3
TRANSITION SERVICE In defining the requirements for the Change Consultancy service, the project team recommends the following 9 stage process (Figure 3 – Service Migration Project Stages) for moving services from an existing data centre to a shared service data centre.
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
35
© PTS Consulting 2010
Figure 3 - Service Migration Project Stages
The stages are focused on ensuring that the required levels of due diligence is carried out on all services prior to migration, and that the services delivered from the shared service are at the appropriate level for delivery to the institution customers. 11.3.1
Stage 1 - Pre-Migration Consultancy and Scoping The first stage of the migration of services should start with a level of awareness and understanding in which services are required to be migrated to the shared service. The NEDCS would meet with the HEI representatives and provide an overview of the project stages that the NEDCS could assist the HEI with, in the migration process. At this stage, a project could be opened and the Customer Relationship Manager (see below) could allocate a project manager and team to assist in the migration process. The process would be aimed at minimising any potential disruption to services and to ensure that the new shared service meets the requirements of the institution in terms of availability, capacity, continuity and performance. As part of this process the Account Manager could also notify the Supplier Manager that a potential new service requirement was being scoped and to make the service provider(s) aware of a potential new service requirement. It would be important at this stage to ensure that the HEI has all the help and support to win their case for the activity they are undertaking. Therefore a number of sub-activities would be supported by the NEDCS, these include: 1. Readiness Assessment – a review of the HEI present status and ability to perform the function. The NEDCS would provide guidance in ensuring the right level and number of resources were available to undertake the project. 2. Winning the case for change – the NEDCS would provide the HEI with support resources to assist the proposer in constructing the business case for the project.
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
36
© PTS Consulting 2010
3.
4. 11.3.2
Developing a high level Milestone Plan – using the experience and expertise from within the NEDCS, an achievable Milestone Plan could be created using experience gained in delivering similar projects at other institutions. Assessing the Human Resource impact and assisting the HEI in planning to manage it.
Stage 2 - Application and Service Mapping and Benchmarking In developing the project for the migration of a service (each service migration would be defined within a separate discreet project plan), it would be important to clearly understand how the service to be migrated is presently configured to work. This is to ensure that the environment for the migrated service is replicated exactly within the new SSDC environment. It is not advisable to concurrently customise and migrate a service into the new environment. The NEDCS would provide access to, and support for, the use of an automated application mapping tool. The tool would sit on the HEI network and listen to the traffic passing across the network, tracking applications and their service requirements (although not guaranteed to capture every service, these tools greatly assist in the capture and development of the service documentation for each migration project). With a map of all the services, there would also be an opportunity to carry out performance benchmarking on key applications. This exercise is used to compare service performance post migration to that pre-migration, and greatly assists in early identification of issues with migrated services.
11.3.3
Stage 3 - Customer Service Requirements With a sound understanding of the services available, and discussing with the IT department users for each service, the NEDCS, in conjunction with the HEI IT group, can meet with service customers to capture and understand their existing and future requirements. The customer’s requirements could be important if services are not to be over-provisioned, thereby costing the HEI more in service subscriptions than is necessary. Identifying customer’s service requirements would also assist the migration project team in identifying effective windows for service migration activity. This may also reduce the overall cost of migrations if services could be migrated during standard office hours, rather than during hours that attract premium rates.
11.3.4
Stage 4 – Service Catalogue Generation In capturing the status of all services from a technology basis, and understanding the customer’s requirements, a catalogue of services initially provided by the HEI IT Department would be validated, and where required updated. The Service Catalogue would identify the services from a customer perspective in the Business Service Catalogue section, and each service would have an individual technology map identifying how the service is configured to work within the Technical Service Catalogue. The Service Catalogue would be the core live document/database that would be used by the project team to identify the service migrations and maintain a record of live services. After each successful migration, the Service Catalogue would be updated under change control to reflect the changes that have occurred as a result of the migration activity.
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
37
© PTS Consulting 2010
11.3.5
Stage 5 – New Service Requirements Generation The service migration project team would utilise the captured data (Customer Requirements and Service Catalogue) to identify the future service requirements from the service provider partners, and would create a service request document used by the NEDCS’s procurement team in sourcing the services required.
11.3.6
Stage 6 – Service Procurement The NEDCS would have already sourced, through an approved procurement framework, data centre and infrastructure capacity. The procurement team would ensure that sufficient capacity and services are allocated to the HEI to cover their existing and predicted capacity requirements. Once these services have been procured, and the service migration team made aware of the intended location for the service, the team can move to the service planning stage.
11.3.7
Stage 7 – Service Migration Planning The migration project manager can develop a programme of works, with a project allocated to each service migration (services may be migrated in service packages, or for larger or priority services, migrated as individual services). Each service would have a definition for the technology requirements, data storage source and target, the support team for the applications (internal and 3rd party), the customers of the service, a time window for the migration activity, and a clearly defined test plan. The NEDCS could assist the HEI in identifying the test plan activity. The testing stages, as a minimum, would be carried out as part of any new installation and would be at least: Functional Testing (pre and post migration) Does the service work with simple tests run by the migration team? These tests and their results would be the same prior and post the migration of the service. The migration team would also plan to test any additional add on services required such as external connectivity, data backups etc. Customer Acceptance Testing The migration project team would plan to work with identified customers and their nominated testers to identify a programme of tests to exercise the service to confirm results and performance are identical or better than the service was prior to the migration activity
11.3.8
Stage 8 – Services Migration Activity The migration project manager would manage the migration of the service from the existing service location to the new shared service data centre. The activities would be conducted using a detailed project task list, and would require successful completion of each task prior to moving to the next stage. Should any element fail a decision would need to be made whether to fix and proceed, or abandon the migration activity and implement the defined back-out plan – leaving the service as it was prior to commencement of the migration. It would be important that the process is conducted under a strict change control process and that all service documentation be updated, and support personnel made aware of any changes to the service during the migration activity. If the service migration was successful the service would be moved into a service warranty period where the migration support team remain on standby should an unexpected subsequent failure
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
38
© PTS Consulting 2010
occur. This is typically for a period of approximately 2 weeks, or sufficient to cover a critical period i.e. month-end processing. 11.3.9
Stage 9 – Retirement and Decommission of old services The final stage, once the change for the migration has been signed off and closed, would be to decommission any retired services. Documentation and support contracts would need to be updated to ensure no unnecessary IT equipment and/or resource was continuing to support a retired service. The aim of the NEDCS would be to provide access to the HEIs for these services, allocating experienced personnel to the HEI’s migration project as requested. These experienced personnel may be full-time employees of the NEDCS, contractors or procured consultancy services capable of assisting the HEI community in completing the migration activity efficiently and effectively.
11.4
SERVICE MANAGEMENT The NEDCS would provide the HEI community with a range of data centre services. These services would also include the procurement and management of the data centre service provider partners, which may include: Data centre facility providers Network providers Infrastructure service provider Software service provider Technical and Business Consultants Contractors The NEDCS would also look to provide an HEI customer relationship management team, to work with the HEI’s in managing the services, and to ensure that the HEIs are receiving the agreed levels of service from the service provider partners. To achieve an efficient management of the service it would be important for two separate support groups to be set up within the NEDCS (Figure 4 – Service Management Activities). The Supplier Relationship Management team would be responsible for managing and monitoring the services provided by the service provider partners, while the Customer Relationship Management team would be responsible for working with the HEIs to understand customer business requirements, converting them into data centre service requirements.
11.4.1
Supplier Relationship Management The Supplier Relationship Managers would be experienced IT Service Managers, understanding the technical delivery requirements of the HEI customers, and the scope of the services contracted for, from the service provider partners. The supplier relationship managers would develop a service strategy in conjunction with the service provider partners aimed at developing services over the term of the service contract.
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
39
© PTS Consulting 2010
Figure 4 - Service Management Activities
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
40
© PTS Consulting 2010
The service strategy could look to continually develop the service for the good of the HEI community, and could cover the four key service areas: 1. Innovation and Learning How could the service be improved through the introduction of new infrastructure or services for the good of the community? What new ways of working could improve the availability, capacity, continuity, performance or value of the service being delivered? 2. Business Process and Quality of Service How could the service provider partner improve the way they deliver their service? Are there areas of service that are not to the expected level of quality? This could cover implementation of services, support of incidents, interaction with the NEDCS, provision of management reporting etc. 3. Customer Relationship How effectively does the service provider partner communicate to the NEDCS at all levels of the business, these may include finance, engineering, account management, directors etc.? Could the service provider partner improve the effectiveness of their communications? 4. Financial How does the service provider partner gain reward for their services? Does the service provider partner still maintain a competitive position in the marketplace for the services provided? What is the accuracy of the billing systems? Are invoices submitted are accurate? Each element of the service provision could be monitored, and a performance report created by the service relationship manager on the service provider partner covering the services delivered. The performance report would require input from different areas of the NEDCS business, but when managed effectively would assist in identifying areas of improvement from the service provider. 11.4.2
Customer Relationship Management The focus on this team is to serve the needs of the HEI community. The customer relationship management team would be experienced HEI personnel who understand the education sector and the needs of the community. The primary objective of this team is to identify business needs and convert them into a service requirement. The customer relationship management team would also discuss service performance addressing services that may not be delivered within agreed levels of service. The customer relationship management team would also provide a regular service review meeting to discuss performance of the NEDCS and the service provider partners, and to discuss ways of improving the service. The Account Team would also address day-to-day, enquires, concerns, and issues raised by the HEI customer community.
11.5
SUBJECT MATTER EXPERTISE The NEDCS would look to source and provide expertise in a number of specific data centre skillsets. These skill-sets could be provided by full-time employees, contractors, consultancies or service provider partners. Providing the knowledge effectively would require identifying where the most efficient source for the information and procuring the delivery of the information for future needs. The types of information can be as diverse as:
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
41
© PTS Consulting 2010
11.5.1
Data Centre Services All areas of data centre services, to assist HEIs with maintaining or upgrading their existing facilities, or developing new ones where they make good business sense. The data centre services subject expertise would cover: Data centre facility management, covering environmental services, electrical supply, cooling technologies, physical security etc. The management of data centre facilities and provision of suitably trained personnel The monitoring and operation of data centre facilities, to include knowledge of data centre toolsets Knowledge of the marketplace, competitors, industry trends, new techniques and technologies
11.5.2
Technology This area covers the technology delivery of the IT infrastructure and the standard architecture design to be used in the delivery of the end-to-end services. Working with the service provider partners, the technology specialists would develop designs and use new technology developments to continually improve the delivery platforms. The technology specialists would specialise in skill sets covering, but not limited to: Networking Network and Infrastructure Security Systems Server platforms High Performance Computing Platforms Server Virtualisation Operating Systems (Windows/Solaris/Linux/BSD) Storage Platforms/Virtualised Storage Infrastructure Management tools Database platforms (Oracle/SQL/Sybase)
11.5.3
Service Management In the delivery of IT service to customers, it is important that services are delivered efficiently and effectively. The NEDCS could employ one of the service management frameworks such as Information Technology Infrastructure Library (ITIL), Control Objectives for Information and related Technologies (CobiT), or Microsoft Operations Framework (MOF). Through the use of industry best practices, as defined within these frameworks, the NEDCS could be sure that the processes they use, and the services they provide, would interface effectively with the customer’s and service provider’s processes. The success of any services provided would be reliant on the effective and efficient implementation of automated processes, and clearly defined and monitored control points. The NEDCS can also provide support and guidance to HEIs in the provision of awareness sessions, training, and the definition and implementation of service management policies and processes. The skill-sets could be provided through a collection of acquired skill-sets provided by internal fulltime staff, contractors, consultancies and/or service provider partners.
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
42
© PTS Consulting 2010
11.5.4
Higher Education Institution With the service specifically targeted initially at the HEI community it would be important that staff are well versed in the culture of the Higher Education community. Staff, contractors, consultancies and service provider partners should be engaged after demonstrating their sound understanding of this community.
11.6
DATA CENTRE ACQUISITION AND SUPPLY As discussed above, the role of the NEDCS would be to define the data centre services requirements (the type of data centre, the type of data centre space, the minimum levels of service, the amount of space, additional services required and network connectivity (JANET)). Then procurement would source the data centre space by either buying from existing commercial service provider partners or existing data centre space within the HEI, or by developing new cost effective data centre space (dependent on an acceptable business case). Using economies of scale, the NEDCS procurement team are able to negotiate and procure services at competitive rates. To assist the procurement team, the Customer Relationship Managers working with the HEIs will need to ensure that future requirements are identified, validated and communicated back to the procurement team for planning. On understanding the customer data centre demands, the NEDCS would acquire sufficient and appropriate data centre space in order to fulfil the community requirements. This would be achieved through working with the customer relationship managers to identify customers and services to be migrated into the new space.
11.7
NEDCS GOVERNANCE In assessing the Governance and Partnership Model to be used in supporting a national data centre service organisation providing services to the wider education community, it was agreed by the SSDC project team that a model based on that used by JANET UK should be implemented. The functions required are as follows: Technology (led by a Chief Technology Officer) Procurement – to lead the purchasing of services from service provider partners using standard procurement frameworks. This is a key role in ensuring that strategic service target requirements are defined within the procurement process Strategic Technologies – this role would look to develop services in conjunction with contracted service provider partners, and assess new technologies that meets the strategic vision for the service and education community Business (led by a Business Director) Sales – the sales operation would look to sell spare capacity so that maximum usage of available procured capacity is achieved. The sales operation would also work with the Customer Relationship Managers to develop and communicate new technology and service opportunities Customer Relationship – this group would own the Higher Education community accounts, ensuring that the customer institutions expectations were met and that satisfaction with services delivered by the NEDCS remains high. The Customer Relationship Managers would address potential new service opportunities and assist in the implementation of new services. The Customer Relationship managers would also work with their institutions to develop awareness programmes promoting the services provided by the NEDCS Operations (led by an Operations Director)
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
43
© PTS Consulting 2010
The Service Desk – is the front office and first point of contact for the NEDCS, they would take calls direct from the service user community and would assist in resolving issues using standard best practice processes Data Centre Operations – responsible for the day to day running of the data centre services. This role could be outsourced to a partner or insourced to a dedicated 24x7 operations team. This service would normally be provided by central Service Operations Centre (SOC) who would have visibility of all data centres and the services provided by them. In the event of an issue, they could can either remotely resolve the issue using management tools or escalate to onsite personnel or third party service provider partners for resolution Supplier Managers – the Supplier Managers would be responsible for the management of service provider partners and the services they provide. The Supplier Managers have an important role in ensuring that services are continually developed in line with the NEDCS strategy through the contracted service provider partners. The Supplier Managers would also work with the service provider partners to review services delivered by the service provider partners and would explore opportunities to increase revenue for the service provider partners through the increased delivery of cost effective innovative services Communications and Support (led by a Communications Director) Publications – responsible for the publication of articles, books etc, on the services and technology provided by the NEDCS. This would be an opportunity for the group to develop thought leading articles and develop best practices and standards for the data centre industry. Events – responsible for developing attendance at Industry Events, Seminars, Webinars, Networking events etc. to publicise the services the NEDCS provides to the community. With a customer base potentially in the many hundreds this would be an important function to ensure the community are fully aware of the benefits the services provided by the NEDCS may bring to individual institutions Association Communications – with many community associations and industry associations linking existing and potential customers, the association network is an important communications medium for generating interest in the services and gaining feedback on existing and future services delivered by the NEDCS Training – not just external training but also internal training of staff is an important function within the NEDCS as the group would look to develop and supply industry leading services. With an opportunity of using internal staff to develop and run external facing training sessions on key subjects covering technology, service management, IT Management, project management, this could be seen as a good revenue generator for the group Web Services – with the drive to develop new innovative ways of delivering services the use of web based solutions would be an important process moving the organisation forwards. Institutions could use the web to: Understand the services the group provides through a sales portal Review services the institution has subscribed to Review invoices and order status Raise incidents and change requests and check their status Review performance dashboards for services subscribed to Access to knowledge portals with service information and technology articles Attend webinar meetings Finance and Administration (led by a Finance Director) SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
44
© PTS Consulting 2010
Office Management – responsible for the NEDCS offices, facilities, office services and systems Finance – responsible for the effective management of the NEDCS finances Human Resources – responsible for the NEDCS’s Human Resources function for permanent, contractor and associate staff ICT – responsible for the supply and support of internal ICT systems to the NEDCS business
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
45
© PTS Consulting 2010
11.8
CONCLUSION In concluding the Services of the NEDCS, the role the NEDCS would be important in defining the future strategy of the Higher Education data centre service, and in the success of delivering highly efficient and cost effective value-based IT services to the HE community. The NEDCS could develop a blueprint for the delivery of IT data centre services that could be used in other areas of the UK public sector through the delivery of: Comprehensive consultancy services designed to support change from initial readiness assessment through winning the case, HR restructuring, service design, implementation, migration and ongoing service management Predictable delivery of IT services through the use of industry best practices Standard delivery of IT systems through simplified IT infrastructure platforms Significant cost savings through economy of scale in the procurement of services Best use of innovation through the use of industry leaders in the provision of infrastructure platforms and industry leading software platforms Effective use of the industry knowledge through the efficient acquisition of Subject Matter Experts (Contractors/Consultants etc.) The NEDCS could supply the much needed one-stop-shop and trusted advisor role that would provide the HEI community with a comfort factor, and opportunity to bridge the general lack of expertise and skill-sets in the provision of cost effective data centre services for the future. The sector would benefit from the availability of a shared service providing: Economies of scale in the procurement of services Experience in the transition of services reducing the average time to migrate services A sector specific organisation, understand the culture, the customers, the services and the expectations A centralised organisation able to engage with key suppliers for and on behalf of an education sector Laying the foundations for the development of IT service delivery into application sharing and the delivery of Software as a Services (SaaS)
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
46
© PTS Consulting 2010
12
NEXT STEPS
12.1
BUSINESS CASE DEVELOPMENT PTS recommends that a Business Case be developed and tested for the creation of an Education Data Centre Service through the implementation of an Incubator Project with the primary target of developing the core framework of a NEDCS service. The Incubator service would also test the basic principles of acquiring best practices from the service provider partner community, and delivering basic data centre services to an identified and invited number of HEI customers, these may be the existing Shared Service Data Centre partners. The services and targets for an Incubator Project could be: The setting up of a customer relationship management process with invited customer institutions The setting up of a supplier management process with relationships with procured service provider partners The definition and implementation of a legal entity and organisation governance structure with key positions filled The identification and engagement of key contractors, consultancies and service provider partners under an incubator project contract The definition and template creation of change processes The delivery of basic services – these may be restricted initially to colocation and disaster recovery services but may be extended as services to include basic infrastructure platform services The benchmarking of existing data centre services within HEIs and demonstrating long term cost savings through utilisation of a shared data centre service Implementation of an efficient procurement process using standard European approved procurement frameworks Identification and invitation to selected customers to utilise the service The definition and implementation up of a suitable support structure (this may be acquired from service provider partners) The definition of a strategy for the Incubator Project and subsequent NEDCS service Demonstration that key targets were met in delivering project stages of the incubator project The objectives of the incubator project would be to demonstrate That the sector is ready to adopt and drive the level of necessary change within their institutions that such a service is economically viable the meeting of key efficiency targets that such a service provides key benefits to the institutions using the service that such a service provides benefit to HEI funders that partnerships with the service provider partners provides benefits to a end-to-end service In conclusion the Incubator Project may be seen as the basis for the continuation into a full national data centre service.
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
47
© PTS Consulting 2010
12.2
INCUBATOR PROJECT The Incubator Project team would work to a clearly defined scope of services and customers. The NEDCS organisation would be set up using staff either from existing institution an assignment basis or through a partnership with a third party organisation able to efficiently provide the initial structure of the NEDCS to operate and develop services for the institutions. The incubator organisation would have a clearly defined role and deliverables with an assured reward (the means for guaranteeing this is yet to be agreed). The early adopters will be closely matched for requirements of service. Through closely matching the early adopter services, the incubator organisation can explore the benefits gained from the sharing of resources, experience and infrastructure services. The incubator organisation would procure experienced specialists to develop the processes and templates required to transition a service into a data centre and would be developed for future use by the NEDCS. The incubator organisation would also procure services from third parties as required, with service charges being passed to the early adopters. The incubator organisation would focus on the financial management of services to compare them to the models developed in the exploratory projects reporting on findings to the project steering board. Such a project should last for approximately two years, which would provide sufficient time to transition the early adopter institution services into the data centre and allow sufficient time for services to stabilise to their new environment. The steering group of the project should review progress made by the incubator organisation and agree on whether the project is viable based on the incubator project and should then move into a live operation at the end of the incubator project phase. As part of project the incubator organisation in conjunction with the project steering group should communicate on a regular basis to the community in general on the progress being made focusing on the lessons learned from the project. Therefore the expected achievements of the Incubator Project would be: Validation of the financial model submitted Demonstration of cultural acceptance of a shared service Demonstration of the benefits of a structured Transition Phase Demonstration that the sharing of services can benefit the community Validation that services can successfully run with negligible impact across the JANET network Validation that service improvements can be achieved through the implementation of best practices (ITIL) Validation that like minded organisation can work together to deliver improvements in the quality of service delivery
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
48
© PTS Consulting 2010
12.3
IMPLEMENTATION ROADMAP The next stage for the setting up of a shared service data centre service is the creation and submission for approval of a Business Case, outlining the opportunities and deliverables for a National Education Data Centre Service. The service should be delivered initially as an Incubator Project to test out the principles of the service and create the framework for the future development of the service (Figure 5 – Incubator Project Roadmap). Presently, concerns have been raised by the community regarding the viability and value to an institution from a shared service, as would be expected from such a cultural change to the future delivery such services to the institutions. The incubator project should aim to address these concerns by demonstrating that a service can be set up, implemented and supported through a structured framework within agreed timescales and to business supporting agreed levels of service, while maintaining cost effectiveness targets. Key stages for the project can be seen as: 1. Development of a Business Case for the next stage of the project 2. Definition of the project organisation structure, achievable service targets and business processes 3. Carry out procurement exercise for initial services from service provider partners 4. Identify and engage with pilot customers 5. Create and modify business processes and tools 6. Commence the delivery of the first services to customers through the initial stages of data centre benchmarking and service mapping 7. Commencement of a migration project of low risk services from early adopter customers 8. Implementation of service support structure (may initially be supplied by service provider) 9. Live services to run for a minimum 6 month stable period post transition phase 10. Presentation of performance reports against set targets 11. Development of next stage strategy and activity plans
Figure 5 - Incubator Project Roadmap
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
49
© PTS Consulting 2010
13
APPENDIX 1 – GLOSSARY OF TERMS Term
Definition
Client
Shared Service Data Centre of Salford and Shared Service Data Centre of Derby
CobiT
Control Objectives for Information and related Technologies – Service Management Framework created by the Information Systems Audit and Control Association (ISACA)
CRM
Customer Relationship Management – Software used to assist organisations to manage the development of relationships with customers
Customer
An organisation that makes use of the shared Service who is not a Public Partner or a Private Partner. Most likely to be another Higher Education Institution.
DMS
Document Management System – a computer system(s) used to track and store electronic documents and/or images of paper documents
DRUPS
Diesel Rotary Uninterruptable Power Supply
EMMAN
East Midlands Metropolitan Area Network. The EMMAN network provides high bandwidth connectivity to the education and research community throughout the East Midlands.
ERP
Enterprise Resource Planning – a computer system used to assist in managing all the functions of a business, company or organisation
FE
Further Education
FEI
Further Education Institutions
HE
Higher Education
HEFCE
Higher Education Funding Council for England
HEI
Higher Education Institution
HEI Partner
A Higher Education Institution member of the Partnership that makes use of the Service, and also determines and oversees how the Service is delivered, managed, and operated. The initial HEI Partners and the Shared Service Data Centre of Salford and the Shared Service Data Centre of Derby.
ISP
Internet Service Provider
ITIL
Information Technology Infrastructure Library. Set of books documenting best practices in the delivery of Information Technology. Owned by the OGC.
JANET
The UK’s education and research network provided by JANET (UK).
KPMG
KPMG is a provider of professional services including audit, tax, financial and risk advisory. KPMG has been engaged as a tax advisor to this current phase of the initiative.
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
50
© PTS Consulting 2010
Term
Definition
Logica
Logica is an IT and business services company. Logica provided consultancy to support the Shared Service Data Centre initiative through the development of a Feasibility Study and Business Plan prior to the current phase of the project.
MOF
Microsoft Operations Framework – Service Management Framework as defined by Microsoft
Partnership
In the context of the Shared Service Data Centre initiative the Partnership is how the Partner organisations would form to provide the Service, and how the delivery of the Service is governed.
PDR
Project Definition Report. This is a document that clearly sets out the responsibilities, scope and the deliverables of the project.
Phase
A project Phase is a significant element of an overall programme that may contain multiple stages.
Project Steering Group
The group with representatives from PTS, the Shared Service Data Centre of Salford, the Shared Service Data Centre of Derby, JANET (UK), and HEFCE, that provides direction, support, advice and guidance to the Project Team.
Project Team
The PTS team actively working to deliver this project.
PUE
Power Utilisation Efficiency – the difference in the amount of power utilised by the whole of the data centre against only that power which is used to power the IT Infrastructure. Usually quoted as number e.g. PUE 2.2
Service
A Service is defined as the offering to the Customers and HEI Partners of a data centre facility.
Service Provider Partner
A provider of data centre and/or infrastructure services who is also is a member of the Partnership.
SHED
Shared High End Data Centre. An initiative funded by HEFCE to provide high-end computing services to the HE community.
SSDC
Shared Service Data Centre – an initiative of HEFCE and the Universities of Derby and Salford, and the primary purpose of this project.
SSVC
Shared Services Virtual Data Centre. An initiative funded by HEFCE to provide virtual/cloud computing services to the HE community. (investigation by YHMAN)
Stage
A sub-element of a project Phase. In the context of this project PTS has defined 6 stages within the Options project phase.
SWOT
Strengths, weaknesses, opportunities, threat
U
The height of the electronic modules standardized as multiples of 1.75 inches (44.45 mm) or one rack unit or "U".
UPS
Uninterruptable Power Supply
VLE
Virtual Learning Environment
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
51
© PTS Consulting 2010
Term
Definition
VPN
Virtual Private Network – A virtual private network (VPN) is a computer network that is implemented in an additional software layer (overlay) on top of an existing larger network for the purpose of creating a private scope of computer communications or providing a secure extension of a private network into an insecure network such as the Internet. (Wikipedia)
YHMAN
Yorkshire and Humberside Metropolitan Area Network. The YHMAN network provides high bandwidth connectivity to the education and research community throughout Yorkshire and Humber. Table 3 - Glossary of Terms
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
52
© PTS Consulting 2010
14
APPENDIX 2 – OBJECTIVE TIERING PROCESS The Objective Tiering Process allows services to be assessed against a pre-defined and agreed set of criteria. The criteria would focus on allowing services to be differentiated through an assessment on two levels – the alignment of the service against the organisations strategy, and, the contribution the service provides to the performance of the business (Figure 6 – Objective Tiering Model).
Services are split into FOUR priorities: 1. Critical 2. Essential 3. Maintenance 4. Utility
Figure 6 - Objective Tiering Model Each service tier would be rated against defined criteria which would be prior agreed. Criteria would vary from organisation to organisation an example of a ‘Criteria Matrix’ is shown below. CRITICAL
ESSENTIAL
MAINTENANCE
UTILITY
Is service key to Business Process?
HIGH
HIGH
HIGH
LOW
Is service Strategic to Business Development?
SIGNIFICANT
SIGNIFICANT
LIMITED
LIMITED
Strategic Alignment
Contribution to Business Performance Experience of Institution Customers (Students)? - OR
SIGNIFICANT
SIGNIFICANT
LIMITED
LIMITED
Institution Reputation potential adverse impact? - OR
SIGNIFICANT
SIGNIFICANT
LIMITED
LIMITED
Financial – potential revenue loss per single incident? - OR
HIGH
MEDIUM
LOW
NONE
Outage would cause operational impact on Institution? - OR
HIGH
MEDIUM
LOW
NONE
SIGNIFICANT
LIMITED
LIMITED
NONE
Outage would cause personal impact/injury?
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
53
© PTS Consulting 2010
Usage of an Objective Tiering Process allows different tiered services to attract different levels of service to be offered and in turn the support structure can be tuned to focus on the support of identified and agreed business critical services. Through the allocation of services to specific objective tier rating process, organisations should see that the greater majority of the services that are supported today will now be rated at a tier 3 or 4 level. PTS has developed through implementation of the tiering system in organisations found the following model as an average rating for each of the tiers (Figure 7 – Modelling Service Allocations).
Figure 7 - Modelling Service Allocations
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
54
© PTS Consulting 2010
15
APPENDIX 3 – FINANCIAL ASSUMPTIONS Notes – the original Logica assumption entries from stage 1 of this project are provided below in Italics. No.
Assumption
Variance
An existing data centre facility or central equipment room (CER) at each institution would be maintained. The shared service data centre would provide either primary or secondary data centre services at Derby, Salford and Leicester 1
2
3
Existing provision at each institution will be maintained. The new shared service data centre will provide secondary business continuity provision at Sheffield Hallam and will provide either primary or secondary data centre services at Salford and Derby. All options provide costs for meeting the projected expansion requirements of the three universities over the forthcoming 10 years and for maintaining existing services All options provide costs for meeting the expansion requirements of the three universities over the coming 10 years and for maintaining existing services. The cost of power will rise at a higher rate than the rate of inflation
None
None
10% above inflation
The cost of power will rise at 5% above the rate of inflation. The cost of co-location rental costs will rise at a higher rate than the rate of inflation.
4
Whilst it is difficult to accurately assess the future rate of increase, all co-location data centre providers will be required to meet future environmental standards and, where additional capacity is required, will need to make capital investment to meet additional M&E costs. These additional costs would be passed on as increases in rental and service charges 10% therefore represents a conservative estimate over the period The cost of co-location rental costs will rise at 10% above inflation. Whilst it is difficult to accurately assess the future rate of increase, all co-location data centre providers will need to meet future environmental standards and, where additional capacity is required, will need to make capital investment to meet additional M&E costs. 10% therefore represents a conservative estimate over the period.
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
55
© PTS Consulting 2010
10% above inflation
No. 5
6
Assumption
Variance
Annual inflation estimate per year (based on average CPI over lat 10 years – data from UK Government National Statistics) The costs for transition to the new data centre and the cost of servers are high level estimates. (see implementation cost estimates) The costs for transition to the new data centre and the cost of servers are high level estimates. The Business Case will need to be updated to reflect more accurate costing during the pre-implementation phase of the programme.
1.8%
None
The cost of ongoing rental of communications links are discounted based on the use of the JANET Network funded by HEFCE. 7
The availability of bandwidth and JANET (UK) acceptance of the use of their network for high availability services will be a key constraint on the use of this service. PTS has obtained general agreement from JANET UK for this use
None
The cost of ongoing rental of communications links are discounted based on the use of SuperJanet connectively funded by HEFCE. The availability of bandwidth and JANET (UK) acceptance of the use of their network for high availability services will be a key constraint on the use of this service. An alternative model with the cost of using commercial communications infrastructure has been included at Annex B. Net VAT impact is included based on the assumption that the shared service company in the sale of good and services to the institutions charges VAT to the institutions and that the shared services company is able to claim Vat back on services purchased 8
(This is assumed based on information provided by KPMG – that the institutions would not be penalised for double VAT on services)
5%
The institutions may be able to reclaim some VAT allowance but this varies from institution to institution – a 5% allowance is applied as an average Net VAT impact is included based on the assumptions stated in Scenario Three of Sub-section 7.2 of the Feasibility Study. Estimated number of staff required at each institution to run 24/7 operation at baseline service level. 9
(It is assumed that to run an effective data centre service on site covering a 24x7 Mon-Fri operation, there is a requirement for 4 shifts of 2 ops plus a Supervisor/Manager) Estimated number of staff required at each institution to run 24/7 operation at baseline service level.(12)
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
56
© PTS Consulting 2010
9
No. 10
11
12
13
13
15
Assumption
Variance
Assumed that Year One is financial year 2010/11, starting 1 August 2010 Assumed that Year One is financial year 2010/11, starting 1 August 2010 (2010/11) Estimated annual on-cost of running existing services at each institution: facilities charges, etc. (Leicester & Salford) Estimated annual on-cost of running existing services at each institution: facilities charges, etc. (£20,000) Estimated annual on-cost of running existing services at each institution: facilities charges, etc. (Derby) Estimated annual on-cost of running existing services at each institution: facilities charges, etc. (£20,000) Shared service data centre will be run on a service basis by service provider partner. Number of staff required by shared service structure to run operation based on the service scope. (24) Total staff required to run existing facilities at each institution and to run and manage the shared service if existing services are maintained. Total staff required to run existing facilities at each institution and to run and manage the shared service if existing services are maintained. (27) Total staff cost per annum (including on costs) – based on average salary times 1.6 (rounded up for calculation) Total staff cost per annum (including oncosts). (£50,000)
2010/11
£20,000
£30,000
None
9
£50,000
16
Staff Cost per day – based on balanced use of internal and external staff
17
Staff Cost per hour – based on 8 working hours per day
£62.50
18
Assumed VAT Rate
17.5%
19
£500
Net Present Value discount Factor (based on Treasury Green Book).
3.5
Net Present Value discount Factor (based on Treasury Green Book. (3.5%).
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
57
© PTS Consulting 2010
No. 20
21
22
Assumption
Variance
Assumed net VAT payable on goods and services procured from 3rd party - after discounting partial exemption. Assumed net vat payable on goods and services procured from 3rd party - after discounting partial exemption. (16.65%) Assumed net vat payable on goods, services and staff procured through Shared Service vehicle - after discounting partial exemption. Assumed net vat payable on goods, services and staff procured through Shared Service vehicle - after discounting partial exemption. (19.5%) Number of non-chassis based servers contained in a rack at a co-located facility. Number of servers contained in a rack at a co-located facility. (5)
16.65%
19.5%
6
23
Number of server chassis installed in a rack (each chassis is approx. 10U) in a co-located facility
3
24
Number of blade servers installed in chassis in a co-located facility
16
25
26
27
28 29
One off cost of installing a rack in a co-located facility.
£1000
One off cost of installing a rack in a co-located facility (£900). Annual cost of running a rack at a co-located facility (excluding power) – Average 3Kw rack Annual cost of running a rack at a co-located facility. (£12,000) Annual cost of running a rack at a co-located facility (excluding power) – Average 10Kw rack
Power - No of Kilowatts used per non-chassis based server per hour
Power – No of Kilowatts used per server chassis per hour
58
£17,000
0.5kW/hr
Power - No of Kwatts used per server per hour (0.5)
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
£12,000
4kW/hr
© PTS Consulting 2010
No. 30 31
Assumption
Variance
Power - Current price per Kilowatt (in pence) for institution power
10p
Power - Current price per Kwatt (in pence) (10p) Power – Current price per Kilowatt (in pence) for colocation power
15p
Power
32
Assumed multiplier to equate Kilowatt per server cost to total power cost for DC, based on use of pre 2005 standards facility. (PUE – Power Utilisation Efficiency Rating) This assumes that full M&E is in place in existing facility.
2.5
Power - Assumed multiplier to equate Kwatt per server cost to total power cost for DC, based on use of pre 2005 standards facility. This assumes that full M&E is in place in existing facility. (3) 33
Assumed PUE – (Power Utilisation Efficiency Rating) of colocation data centre
1.6
34
Assumed PUE – (Power Utilisation Efficiency Rating) of new build data centre
1.3
35
Assumed PUE – (Power Utilisation Efficiency Rating) of refurbished build data centre
1.6
HEFCE and other grants available to pump prime shared service provision. Assumed these will not be repayable. 36
37
38
39
HEFCE and other grants available to pump prime shared service provision. Assumed these will not be repayable. (£3,000,000) Assumed Capital Cost (for Option A) of bringing existing accommodation to baseline level - Salford. Assumed Capital Cost (for Option A) of bringing existing accommodation to baseline level – Salford. (£500,000) Assumed Capital Cost (for Option A) of bringing existing accommodation to baseline level - Derby. Assumed Capital Cost (for Option A) of bringing existing accommodation to baseline level - Derby. (£1,000,000) Assumed Capital Cost (for Option A) of bringing existing accommodation to baseline level - Leicester. Assumed Capital Cost (for Option A) of bringing existing accommodation to baseline level - Sheffield Hallam. (£250,000)
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
59
© PTS Consulting 2010
None
£500,000
£1,400,000
£1,000,000
No. 40
41
Assumption
Variance
Assumed annual revenue cost (for Option A) of bringing existing accommodation to baseline level - Salford. Assumed annual revenue cost (for Option A) of bringing existing accommodation to baseline level - Salford. (£250,000) Assumed annual revenue cost (for Option A) of bringing existing accommodation to baseline level - Derby. Assumed annual revenue cost (for Option A) of bringing existing accommodation to baseline level - Derby. (£100,000)
£50,000
£100,000
Assumed annual revenue cost (for Option A) of bringing existing accommodation to baseline level – Leicester 42
43
44
45
46
Assumed annual revenue cost (for Option A) of bringing existing accommodation to baseline level - Sheffield Hallam. (£10,000) Assumed additional capital cost (for Option A) of enabling existing accommodation to meet expansion requirements Salford. Assumed additional capital cost (for Option A) of enabling existing accommodation to meet expansion requirements Salford. (£1,500,000) Assumed additional capital cost (for Option A) of enabling existing accommodation to meet expansion requirements Derby. Assumed additional capital cost (for Option A) of enabling existing accommodation to meet expansion requirements Derby. (£4,500,000) Assumed additional capital cost (for Option A) of enabling existing accommodation to meet expansion requirements – Leicester. Assumed additional capital cost (for Option A) of enabling existing accommodation to meet expansion requirements – Sheffield Hallam. (£4,500,000) Assumed annual revenue cost (for Option A) of maintaining existing accommodation based on expanding requirements - Salford. Assumed annual revenue cost (for Option A) of maintaining existing accommodation based on expanding requirements Salford. (£150,000)
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
60
© PTS Consulting 2010
£100,000
£1,500,000
£1,500,000
£4,500,000
£150,000
No.
47
48
Assumption
Variance
Assumed annual revenue cost (for Option A) of maintaining existing accommodation based on expanding requirements - Derby. Assumed annual revenue cost (for Option A) of maintaining existing accommodation based on expanding requirements Derby. (£450,000) Assumed annual revenue cost (for Option A) of maintaining existing accommodation based on expanding requirements - Leicester. Assumed annual revenue cost (for Option A) of maintaining existing accommodation based on expanding requirements Sheffield Hallam. (£450,000)
£150,000
£450,000
Annual savings generated from re-use of existing accommodation based on transition of existing services- Salford 49
Annual savings generated from re-use of existing accommodation based on transition of existing services- Salford (£20,000)
£20,000
Annual savings generated from re-use of existing accommodation based on transition of existing services- Derby 50
Annual savings generated from re-use of existing accommodation based on transition of existing services- Derby (£10,000)
£20,000
Annual savings generated from re-use of existing accommodation based on transition of existing services- Leicester 51
52
Annual savings generated from re-use of existing accommodation based on transition of existing services- Sheffield Hallam (£20,000) Annual savings generated from improved services to users as a result of more capacity, better standards, zero outages, improved SLAs, etc. - included in all Options except A - Salford. Annual savings generated from improved services to users as a result of more capacity, better standards, zero outtages, improved SLAs, etc. - included in all Options except A - Salford. (£20,000)
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
61
© PTS Consulting 2010
£40,000
£20,000
No.
53
54
55
56
57
58
Assumption
Variance
Annual savings generated from improved services to users as a result of more capacity, better standards, zero outages, improved SLAs, etc. - included in all Options except A - Derby. Annual savings generated from improved services to users as a result of more capacity, better standards, zero outtages, improved SLAs, etc. - included in all Options except A - Derby. (£100,000) Annual savings generated from improved services to users as a result of more capacity, better standards, zero outages, improved SLAs, etc. - included in all Options except A - Leicester. Annual savings generated from improved services to users as a result of more capacity, better standards, zero outtages, improved SLAs, etc. - included in all Options except A - Sheffield Hallam. (£20,000) Assumed cost for each institution of migrating existing services to shared service data centre. Will need validation in pilot phase. Assumed cost for each institution of migrating existing server estate to new facility. Will need validation in pilot phase. (£300,000) Assumed cost per server purchase and installation both to meet additional expansion requirements and also if existing DC facilities are migrated. Will need validation in pilot phase. Assumed cost per server purchase and installation both to meet additional expansion requirements and also if existing DC facilities are migrated. Will need validation in pilot phase. (£0) Cost of Borrowing (figure used by finance department when conducting investment appraisals). Cost of Borrowing (figure used by finance department when conducting investment appraisals). (7%) Deflator (figure used by finance department when conducting investment appraisals). Deflator (figure used by finance department when conducting investment appraisals). (3%)
£100,000
£40,000
Dependent on each institution
£0
7.0%
3.0%
59
Number of servers / server instances presently used in Derby
520
60
Number of servers / server instances presently used in Leicester
670
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
62
© PTS Consulting 2010
No.
Assumption
Variance
61
Number of servers / server instances presently used in Salford
300
62
Estimated Number of services presently in use in Derby
200
63
Estimated Number of services presently in use in Leicester
500
64
Estimated Number of services presently in use in Salford
200
65
Existing data centre floor-space – Salford
2002sq foot
66
Existing data centre floor-space – Derby
2044sq foot
67
Existing data centre floor-space – Leicester
8363sq foot
68
Average Cabinet Power consumption within institutions (Derby / Leicester / Salford) – Kw/hr – Low Virtualisation
69
Number of days in a year
70
Number of hours within a year (365x24)
71
Existing Cabinet count for Leicester
120
72
Existing Cabinet count for Derby
65
73
Existing Cabinet count for Salford
60
74
Annual Cabinet Growth per year - Leicester
5
75
Annual Cabinet Growth per year - Derby
3
76
Annual Cabinet Growth per year - Salford
5
77
Aggressive Virtualisation Cabinet Count estimate – day one - Leicester
50
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
3kW/hr 365
63
8760hrs
© PTS Consulting 2010
No.
Assumption
Variance
78
Aggressive Virtualisation Cabinet Count estimate – day one - Derby
15
79
Aggressive Virtualisation Cabinet Count estimate – day one - Salford
15
80
Average cabinet power consumption within institutions (Derby / Leicester / Salford) kW/hr – Aggressive Virtualisation
9kW/hr
81
Average Cabinet Power consumption within institutions (Leicester) - Kw/hr - Aggressive Virtualisation plus mix of equipment
6kW/hr
81
Aggressive Virtualisation annual cabinet growth - Leicester
3
82
Aggressive Virtualisation annual cabinet growth - Derby
1
83
Aggressive Virtualisation annual cabinet growth - Salford
1
84
Number of Services per Server Chassis (Aggressive Virtualisation( (16 Servers – 60 services per Chassis)
60
SSDC Final Options Analysis Report J:\Central\Shared Service Data Centre
64
© PTS Consulting 2010
