Linux-HA

The Linux-HA (High-Availability Linux) project provides a high-availability (clustering) solution for Linux, FreeBSD, OpenBSD, Solaris and Mac OS X which promotes reliability, availability, and serviceability (RAS).

The project's main software product is Heartbeat, a GPL-licensed portable cluster management program for high-availability clustering. Its most important features are:

* no fixed maximum number of nodes - Heartbeat can be used to build large clusters as well as very simple ones
* resource monitoring: resources can be automatically restarted or moved to another node on failure
* fencing mechanism to remove failed nodes from the cluster
* sophisticated policy-based resource management, resource inter-dependencies and constraints
* time-based rules allow for different policies depending on time
* several resource scripts (for Apache, DB2, Oracle, PostgreSQL etc.) included
* GUI for configuring, controlling and monitoring resources and nodes

History

The project originated from a mailing list started in November 1997.Eventually Harald Milz wrote an odd sort of Linux-HA HOWTO. Unlikemost HOWTOs, this one was not on how to configure or use existing software, it wasa collection of HA techniques which one could use if one were to write HA software for Linux.

Alan Robertson was inspired by this description and thought that he could perhaps write some of thesoftware for the project to act as a sort of initial seed crystal to help jump start the project.He got this initial software running on [http://lists.linux-ha.org/pipermail/linux-ha/1998-March/000076.html 18 March, 1998] . He created the first web site for the project on [http://lists.linux-ha.org/pipermail/linux-ha/1998-October/001274.html 19 October, 1998] ,and the first version of the software was released on [http://lists.linux-ha.org/pipermail/linux-ha/1998-November/002335.html 15 November, 1998] .The first production customer of the software was Rudy Pawul of ISO-NE. The ISO-NE web sitewent into production in the second half of 1999.

At this point, the project was limited to two nodes and very simple takeover semantics, andno resource monitoring.

This was cured with version 2 of the software, which added n-node clusters, resource monitoring,dependencies, and policies. Version 2.0.0 came out in [http://lists.linux-ha.org/pipermail/linux-ha/2005-July/015449.html 29 July, 2005] .This release represented another important milestone as it was the first version where very large contributions(in terms of code size) were made by the Linux-HA community at large.This series of releases brought the project to a level of feature parity-or-superioritywith respect to commercial HA software.

See also

* Carrier Grade Linux
* High-availability cluster

External links

* [http://www.linux-ha.org/ Linux-HA website]
* [http://linux-ha.org/FactSheetv2 Linux-HA fact sheet - provides a detailed list of features]
* [http://wiki.linux-ha.org/ Linux-HA Wiki]
* [http://lists.linux-ha.org/pipermail/linux-ha/ archives of Linux-HA's main mailing list]

References

* Gregory F. Pfister: "In Search of Clusters", 2nd ed., Prentice Hall, ISBN 0-13-899709-8
* Evan Marcus, Hal Stern: "Blueprints for High Availability", 2nd ed., John Wiley & Sons, ISBN 0-471-43026-9