Failure as process

Abstract

Disaster, loss, and failure preoccupy the minds of many digital preservation professionals, and yet, despite the prominence of digital disaster planning guidelines which seem to anticipate failure, there is limited discussion of experience with preservation system or network failures, which are often framed as inevitable in digital preservation. Despite this framing, negative perceptions of failure influence the digital preservation discourse by associating failure with poor planning, unreliability, and untrustworthiness on the part of institutions. This article will interrogate the issue of failure within the digital preservation field and consider the need for more conversations around network failure and recovery. The authors will argue that failure is part of the process of digital preservation and more honest conversations around this topic will contribute to the practice of openness and transparency within the digital preservation community. To illustrate these issues, the authors will discuss the actual hardware failures experienced by the MetaArchive Cooperative, a community-based distributed digital preservation network, and how the Cooperative’s utilization of the LOCKSS software allowed it to recover from those failures. Additionally, the lessons learned and resulting changes the Cooperative made to technical infrastructure, hardware diversity, policies and procedures will be shared.