[Discuss] Redundant array of inexpensive servers: clustering?

[feenberg at nber.org (Daniel Feenberg)]

On Mon, 31 Mar 2014, Rich Braun wrote:

> Edward Ned Harvey wrote:
>> Hehhehe - No.  The goal is mash the power button, with the results described
>> above, while using only 2 servers and free software.   ;-)
>
> Well, if the free or low-cost software existed to make it work well, I'd
> eagerly pony up to pay the electric bill running a third server.  My point
> is--there have been essentially no advances in the software to operate this,
> commercial /or/ free, during the 2000s by anyone other than the paid staff at
> companies like AWS, Google, and Microsoft.  Even VMware's solution is about a
> decade old by now, and its customers (like my employer) are fleeing en masse
> from private VMware clusters to AWS despite the steep price.
>
> -rich
>

Isn't the lesson here that if you can't change the client, this is really 
hard, but if you can get cooperation from the client, then it is 
reasonably easy?  For example, an HA webserver with all the bells and 
whistles might be quite hard to build, but a client that stored state in 
cookies and knew to try www2.example.com when www1.example.com failed 
would be pretty straightforward. I don't know why client-side HA features 
have never shown up in standards since DNS was defined, but they haven't.

Daniel Feenberg