Large database with High Availability - how to do it?

[Daniel Ek]

I really thought that MySQL tried to produce MySQL cluster for common 
hardware and with the design principle; more nodes with common hardware, 
insted of few nodes with extreme hardware. That to get higher 
availability (Much like the GFS (Google File System)). If the memory 
usage (RAM) as I have seen being recommended (double the actual database 
size) for MySQL cluster in this blog: 
http://jeremy.zawodny.com/blog/archives/001868.html, I will probably 
never work for a company that can afford that kind of "super computers".

I today have a database using 20GB storage, which actually means that I 
should have at least 30GB (40GB recommended) ram using Mysql cluster. I 
am today running 2 machines with 2 x XEON 3,0GHZ and 2GB ram with 
replication. What I would like to add is more in that kind of range.

Is there anyone that have a comment regarding how MySQL cluster acts 
with common hardware (if at all possible) or knows other solutions that 
you can make high availability and share the load to the DB servers 
please give me some suggestions.

Cheers
Daniel
PS.
For those who would like to know more about the GFS: 
(http://www.google.se/search?q=cache:m0TMQYgIlIoJ:www.cs.rochester.edu/sosp2003/papers/p125-ghemawat.pdf+google+file+system&hl=sv)

--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:    http://lists.mysql.com/[EMAIL PROTECTED]