In May 1998 a paper entitled ``Creating Redundant Linux Servers'' was presented at Linux Expo[1]. In November the IP Address takeover package detailed in the paper was released as ``fake''. This was arguably the first High Availability software released for Linux. During the same year Alan Robertson started a Linux High Availability page, focusing on the Linux High Availability HOWTO[6].
In the short time that has passed since then a myriad of closed and open source high availability solutions have become available for Linux. Fake has largely been superseded by Heartbeat. ``Alan Robertson's HA Page'', now known as ``Linux High Availability'' is still a focus for much of the activity around high availability for Linux and can be found at www.linux-ha.org. High availability under Linux is no longer limited to IP Address Takeover as both intelligent DNS1 and Layer 4 Switching solutions become available.
The focus of this paper is to examine some of the technologies that are currently available under Linux and how these can be used to create Web Farms. Web farms provide an interesting application of these technologies and is an area where Linux is becoming more and more accepted. Beyond this the paper will examine some of the challenges facing Linux high availability and in particular the directions that open source efforts are taking in order to address these problems.