Service Level Agreements for Storage and Preservation

You are here

Technical Report
Stephen Phillips
Not Applicable
Service Level Agreements

This technical report is a product of the PrestoPRIME Project (ID3.4.1.), the major project on digital preservation in the audiovisual sector in Europe.

Digital audiovisual material can be stored on a variety of hardware types, from spinning discs in a server to tapes on shelves. Regardless of the hardware format, some company or department has the task of storing, preserving and providing access to the data. This document looks at preservation services, whether out-sourced or in-house and the considerations that are important when defining the service level agreement (SLA) with the service provider. Defining the service level agreement is just one task of many that must be carried out when defining the relationship between a producer or consumer and an archive. This document contains two major new pieces of work:

1. A survey of audiovisual service providers, investigating what they consider the important considerations are when determining whether to trust a service provider.

2. A detailed proposal of terms to be included in a service level agreement with a preservation service provider.
Recognising that a service level agreement cannot be sufficient in itself to persuade a content producer or consumer to trust anyone with some data, we conducted a survey amongst professionals from the audiovisual service provider community to see what they thought was most important for convincing someone to use a preservation service. We found that 39% of the respondents were aware of TRAC and that, regardless of prior knowledge, 89% would take into account such an audit certificate when determining whether or not to trust a service provider. We therefore conclude that an audit of a preservation service provider by an independent expert auditor using a scheme such as TRAC is a practical solution to the problem of how to convince someone to use the service.

The majority of this document deals with a proposal of suitable terms to be used in a service level agreement with a preservation service provider.  Firstly a vocabulary is defined to express the necessary concepts and a discussion of monitoring and management techniques and architectures follows. A framework for gathering appropriate SLA terms is presented and has been used in making the proposal of SLA terms presented here in one chapter and an extensive appendix.
Finally, this document is grounded in the reality of the existing relationship between the Dutch organisations Sound and Vision and Technicolor. Sound and Vision provides access to 700,000 hours of Dutch television, radio, music and film and uses the services of Technicolor to store and preserve this content. The two organisations have an existing service level agreement, some details of which are presented here. The commentary on this service provider/client relationship informs us that whereas some business relationship may be quite detached, in the case of a preservation service provider the client must be invited in to understand commercially sensitive information that would not normally be shared. This provides the final reassurance and confidence to use a service provider as well as the necessary information to draw up an exit strategy to guarantee the long-term safekeeping of the material.

The digest of this report can be found here. It is recommended that this document be read in conjunction with the related PrestoCentre documents on threats to mass storage and on storage services to provide a broad round up of information on all the topics relevant to choosing and designing a storage service.