Description:
Why is redo so important ? How does recovery actually work ? What is "a
checkpoint" ? Does the database writer write uncommitted data ? How does
the log writer cope with a very long transaction ? All these questions, and
many more, will be answered through this painless description of how redo
works.
This presentation explains what redo is, why it matters, and how it allows
a crashed database to recover. Topics include redo generation, commits,
cleanout, checkpoints of many types, and the recovery mechanism.
Go to top