Instant Recovery

Instant Recovery logo

The recovery component of a database system takes up the hard task of ensuring atomicity and durability of every modification made to a database (the "A" and "D" of ACID transactions).

While the system is running, the logging component keeps track of what happens with the data, collecting information necessary to undo and redo modifications.


When a failure occurs, recovery kicks in and restores the data to the most recent transactionally consistent state, based on the information it finds in the logs.


The problem we try to solve is that recovery is a painful process, which leaves the system unavailable for hours or even days. Our goal is to design simple algorithms that greatly improve the availability and robustness of the system, without noticeable impacts on runtime performance.

Open source repositories

Since October 2014, our research has moved into the Zero project, which is a fork of Shore-MT initially implemented at HP Labs. Our experiment code and the Shore-Kits framework are also being migrated into a new application called Zapps.

Our code for Single-Pass Restore (SPR), which was published on the BTW 2015 conference and based on Shore-MT, is currently available in BitBucket, in the following pages:



Caetano Sauer, Goetz Graefe and Theo Härder
Instant restore after a media failure
In: Proc. ADBIS 2017, LNCS 10509, pp. 311-325
September 2017
Caetano Sauer
Modern techniques for transaction-oriented database recovery
Ph.D. Thesis, University of Kaiserslautern, Verlag Dr. Hut, München, August 2017
Caetano Sauer, Gilson Souza, Goetz Graefe and Theo Härder
Come and crash our database! – Instant recovery in action
In: Proc. EDBT 2017, pp. 554-557
March 2017
Caetano Sauer, Goetz Graefe and Theo Härder
Instant restore after a media failure
February 2017
Caetano Sauer
Exploring a continuum between main-memory and disk-oriented OLTP systems
CIDR Abstracts


Stefan Hemmer
Key-Value Storage Engines in Relational Databases
Master's Thesis, TU Kaiserslautern, December 2016
Caetano Sauer, Lucas Lersch, Theo Härder and Goetz Graefe
Update propagation strategies for high-performance OLTP (best paper award)
In: Proc. ADBIS 2016, LNCS 9809, pp. 152-165
Springer-Verlag, August 2016
Lucas Lersch, Caetano Sauer and Theo Härder
Decoupling Persistence Services from DBMS Buffer Management
In: Proc. Workshop "Grundlagen von Datenbanken" (GvDB2016)
May 2016


Lucas Lersch
Decoupled Propagation for DBMS Architectures
Master's Thesis, TU Kaiserslautern, December 2015
Goetz Graefe, Caetano Sauer, Wey Guy and Theo Härder
Instant recovery with write-ahead logging
Datenbank-Spektrum 15(3), pp. 235-239
November 2015
Caetano Sauer, Goetz Graefe and Theo Härder
Single-pass restore after a media failure
In: Proc. BTW, LNI 241, Hamburg, pp. 217-236
March 2015


Goetz Graefe, Wey Guy and Caetano Sauer
Instant Recovery with Write-Ahead Logging: Page Repair, System Restart, and Media Restore
Synthesis Lectures on Data Management
Morgan & Claypool Publishers, November 2014
Previous | 1, 2 | Next
Export as: