The terms “Server Load Balancing” and “Clustering” are both somewhat generic terms that have come to mean very specific things. Server Load Balancing refers to appliance-style network devices, and clustering is traditionally a software solution sitting on a server, both of which accomplish about the same thing. One term could be used to describe what the other one does, and vice versa. Still, that’s what they go by traditionally in the IT environment. The English language is replete with such oddities.
Which brings me to my point: Clustering, for the most part, lost. Appliance-style server load balancers won out handily, leaving very little for clustering in the market. There’s only one clustering vendor I know of, Resonate, that even offers a product anymore, and I’m not convinced they’re selling a significant number of units.
So why did appliance based load balancers win out? They’re just a helluva lot easier to deploy, that’s why. Because they operate on the network level, they avoid most of the pitfals of operating system, operating system version/patch level, web server and sever system version/patch levels.
Resonate supports a wide variety of operating systems and version levels, but it’s still a limited number compared to a network appliance-based solution (which is anything, including a coffee maker running a web server).
So what’s my point? None really, other than perhaps: In your face, clustering.
