Cassandra Archiver for CSS

Table of Content

About the Cassandra Archiver

The Cassandra Archiver for Control System Studio is a set of plugins for CSS / EPICS, that use an Apache Cassandra database to archive control system values, providing outstanding performance compared to traditional archiving to a relational-database management-system.


An extensive manual, explaining the concepts of the Cassandra Archiver, a configuration guide and a step-by-step installation guide is included in the binary distribution and also available for online view.


This program and the accompanying materials are made available under the terms of the Eclipse Public License v1.0. By downloading the software you agree to be bound the the terms of this license. In particular, you have to agree to sections five (“NO WARRANTY”) and six (“DISCLAIMER OF LIABILITY”) before downloading the software.

Some files in the binary distribution have different copyrights and licenses. Refer to the information provided in the package for details about those copyrights and licenses.

The Cassandra Archiver for CSS has been replaced by the Cassandra PV Archiver. New users should use the Cassandra PV Archiver. Existing users should consider migrating to the Cassandra PV Archiver. This page is kept online so that existing users can access the documentation and old releases of the Cassandra Archiver for CSS.

Release 2.3.1 (June 6th, 2016):
Fixed a bug in the reader part of the archiver. This bug could cause many samples before the actually queried period to be returned instead of only returning a single sample like intended. As a side effect, the handling of exceptions that might occur while reading samples from the archive was improved slightly. This release only makes changes to the reader part. The archive engine, configuration utility, and clean-up tool remain unchanged.

Release 2.3.0 (October 20th, 2014):
This release brings a number of minor improvements: It bundles a patched version of the CSS archive engine that correctly reconnects after the connection to the database has been lost. It also has improvements in the clean-up tool, that make the clean-up process less likely to fail if interrupted. Even if the clean-up process is interrupted, it is applied at least partially. Finally, there are a number of improvements in the sample retention process and the archive reader, that help avoid burdening the Cassandra database with many tombstones for deleted samples.

Release 2.2.0 (July 28th, 2013):
This release brings performance optimizations in the sample compressor. When using compressed samples, the new code will reduce the number of database requests needed significantly, thus reducing the CPU and I/O load. The format of the stored data has not changed with this release, thus it is fully compatible with the 2.1.0 release.

Release 2.1.0 (March 13th, 2013):
This release brings support for Apache Cassandra 1.2, which offers some very useful new features (e.g. virtual nodes). This release also introduces a revised serialization format for sample values, that is more space efficient for integer values. As a consequence older releases cannot read data written by this release of the Cassandra Archiver, however this version can still read the old format, allowing for an upgrade without any conversion needed.

Release 2.0.1 (March 10th, 2013):
This release fixes a bug that appeared when the interval configured for samples was very long (approximately more than 10 hours). Usually, such an interval is only used with compressed samples, which limited the scope of this bug.

Release 2.0.0 (January 7th, 2013):
This release uses a new structure for the Cassandra database, bringing a major boost in performance. It also fixes some bugs and brings general stability improvements. This release of the Cassandra Archiver needs Apache Cassandra 1.1.x.

Release 1.2.0 (September 17th, 2012):
This release brings support for multiple compression threads. This can help if the compression does not "catch up" with the new data. It also improves the queuing mechanism used for compression to be more memory efficient.

Release 1.1.0 (January 28th, 2012):
This release uses an improved compression algorithm for calculating compressed samples. If a suitable compression level with a shorter compression period is available, the samples from this compression level are used instead of using raw samples. This increases the throughput of the samples compressor significantly.

Release 1.0.1 (January 24th, 2012):
Bugfix release: Fixed Cassandra authentication problem in archive engine. Fixed infinite loop in archive reader (also affected sample compression). Included plugin_customization.ini in archive engine product. Fixed a bug in the sample cache used by the sample compressor.

Release 1.0.0 (January 20th, 2012):
First release of the Cassandra Archiver.


You can contact the author of this software, Sebastian Marsching, at cassandra-archiver-maintainer removethis <at> oss <dot> aquenos <dot> com.

Visit other projects on

Imprint | Privacy Policy

© 2012-2014 aquenos GmbH