Definition, Detection, and Recovery of Single-Page Failures, a Fourth Class of Database Failures

TL;DR

This paper introduces single-page failures as a new failure class in databases, along with a recovery method using a page recovery index that allows fast detection and recovery without aborting transactions.

Contribution

It defines single-page failures as a distinct failure class and proposes a novel recovery structure, the page recovery index, enabling efficient detection and recovery.

Findings

Single-page failures are a new failure class in databases.

The proposed page recovery index enables fast detection and recovery.

Recovery can be performed with minimal transaction delay.

Abstract

The three traditional failure classes are system, media, and transaction failures. Sometimes, however, modern storage exhibits failures that differ from all of those. In order to capture and describe such cases, single-page failures are introduced as a fourth failure class. This class encompasses all failures to read a data page correctly and with plausible contents despite all correction attempts in lower system levels. Efficient recovery seems to require a new data structure called the page recovery index. Its transactional maintenance can be accomplished writing the same number of log records as today's efficient implementations of logging and recovery. Detection and recovery of a single-page failure can be sufficiently fast that the affected data access is merely delayed, without the need to abort the transaction.