A Fault Model for Upgrades in Distributed Systems

Recent studies, and a large body of anecdotal evidence, suggest that upgrades are unreliable and often end in failure, causing downtime and data-loss. While this is sometimes due to software defects in the new version, most upgrade failures are the result of faults in the upgrade procedure, such as broken dependencies. In this paper, the authors present data on upgrade failures from three independent sources - a user study, a survey and a field study - and, through statistical cluster analysis; they construct a novel fault model for upgrades in distributed systems.