Re: [gentoo-server] MySQL-4.1 and <- Max

[Kashani]

On Thu, 28 Oct 2004, Billy wrote:

> http://mysql.mirror.nedlinux.nl/tech-resources/articles/storage-engine/part_2.html
>
> While nodes keep their data entirely in memory for maximum performance,
> the durability requirement of ACID transactions is maintained by the
> cluster making sure that at least two nodes in each group have the a
> copy of the data. On-disk "redo" logs are maintained for recovery
> purposes., and each node requires sufficient disk space to store its
> entire fragment for cold starts.
>
> [This tells me that each node doesn't need to store the entire database,
> just the portion of the database that's responsible for. The more nodes
> you have, the less memory each node requires.]
>
> Because MySQL Cluster is main-memory based it is extremely fast. MySQL
> Cluster has good scalability by adding more memory and/or more nodes to
> a configuration.
>

	Alright, I'm still having trouble wrapping my head around this, 
but it's starting to make sense. Supposing I have a 4 GB database and two 
computers with two nodes a piece, how much RAM do I need in each computer? 
If I'm understanding this correctly I'd need to allocate 2GB per node or 
4GB per server. Assuming that's correct then where this would start making 
sense is at say four servers with 2 nodes a piece. My 4GB database becomes 
four partitions * 2 for ACID compliance or 8GB across the 4 servers. So 
scalibility sucks at < 3 servers like all n+1 solutions.

Is this right? 

kashani