Percona Server moved to https://jira.percona.com/projects/PS

Overview
Code
Bugs
Blueprints
Translations
Answers

Bug #1625466
Comment #1

Comment 1 for bug 1625466

Revision history for this message

Laurynas Biveinis (laurynas-biveinis) wrote on 2016-12-20:

This is caused by the INNODB_CHANGED_PAGES iterator reading one more page past the query range and then stopping the iteration by comparing its start LSN to range end LSN. If that page happens to be corrupted, it would needlessly fail the query with a "Can't find a system record".