The Ganymed Project
Databases are facing important challenges in terms of
functionality, scalability, and extensibility. This is
particularly true for many new applications: genomics, sensor
networks, mobile objects, etc.
From our experience with data
repositories for astrophysics (HEDC), we find that a
big obstacle in using databases within the information
infrastructure for these novel applications is that it is
difficult to extend a database engine. Vendors require long cycles
to incorporate new functionality. It is also not clear that there
is enough of a market demand for most of the functionality
required in scientific applications. Also, many of these new
applications demand a degree of scalability difficult to provide
with centralized engines.
Replication is a common solution to scalability problems but it involves difficult trade-offs.
The most difficult one is the choice between scalability and
consistency. Commercial systems tend to give up consistency. Research
solutions typically either offer a compromise (limited scalability
in exchange for consistency) or impose limitations on the data
schema and the workload.
To address these problems, we are exploring new architectures extending and opening up
database engines. The idea is to combine the power and reliability
of commercial engines with the openness, low cost, and reduced
complexity of open source engines.
Ganymed is our lightweight, experimental platform which
we use in our studies.
People
Publications
- Christian Plattner, Gustavo Alonso, M. Tamer Özsu:
DBFarm: A Scalable Cluster for Multiple Databases.
In: Proc. of the 7th ACM/IFIP/USENIX International Middleware Conference, Melbourne, Australia, November 27 - December 1, 2006.
- Christian Plattner, Gustavo Alonso, M. Tamer Özsu:
Extending DBMSs with Satellite Databases.
Accepted for Publication in the VLDB Journal.
- Christian Plattner, Andreas Wapf, Gustavo Alonso:
Searching in Time.
Proceedings of SIGMOD '06, 2006 ACM SIGMOD International Conference on Management of Data, June 27-29, 2006, Chicago, Illinois, USA.
- Christian Plattner, Gustavo Alonso:
Ganymed: Scalable Replication for Transactional Web Applications.
Proceedings of the 5th ACM/IFIP/USENIX International Middleware Conference, Toronto, Canada, October 18-22, 2004.
- Christian Plattner, Gustavo Alonso:
Ganymed: Scalable and Flexible Replication.
IEEE Data Engineering Bulletin. Vol. 27, No. 2, June 2004.
Download
You can download a copy of our system here: ganymed-dist-1.3.zip. The zip includes the source code for the central parts of the Ganymed prototype system. There is also a manual that describes how to setup the system.
Students that have contributed
- Summer 2006: Tina Körner, VT100 Terminal Emulation for Ganymed SSH-2, semester work.
- Spring 2004: Stefan Nägeli, Ganymed Manager Console, semester work.
- Summer 2004: Mario Tadey, Efficient Serialization of Java Resultsets, semester work.
- Summer 2004: Roger Rüegg, Improving the Ganymed System, master thesis.
-
- Winter 2004: Thomas Gloor, A Java based TPC-C Implementation, semester work.
- Winter 2005: Marcel Bihr, Ganymed Statistics Module, semester work.
- Winter 2005: Christoph Schwank, A Graphical Transaction Analyzer, semester work.
- Winter 2005: Franziska Meyer, TARM: Transparent and Autonomous Replication with MySQL, master thesis.
Additional work has been done by Andreas Wapf and Roger Jäggi.
|