Hi Stewart --
Thanks for the swift reply!
> The *entire* database (this can be spread around your storage nodes
> though).
Please explain what you mean by "spread around your storage nodes". Do
you mean that if I calculate that I need 1 gig of RAM, that I could
allocate 250 megs of ram on each of four cluster nodes?
> The documentation has some formulas that should help you work out
> roughly how much memory your database will consume.
Are you referring to the text which explains that I must total up the
length of each column, and then add about 128k per record? Or is there
something else you could refer me to?
On Alex Davies' excellent web page
(http://www.davz.net/static/howto/mysqlcluster) he says the following:
---START--->
Use the following formula to work out the amount of RAM that you need on
each storage node:
(Size of database * NumberofReplicas * 1.1) / Number of storage nodes
NumberofReplicas is set to two by default. You can change it in
config.ini if you want. So for example to run a 4GB database you need
just under 9GB of RAM on each storage node. For the API nodes and
managment nodes you dont need much RAM at all.
<---END---
This certainly is easier than adding up all the minutia. IOW, on disk I
have a database which takes 27 megs. I have two nodes on my cluster
(not counting the management node). So:
27 * 2 * 1.1 = 59.4 megs
Does that sound right to you?
Thanks,
-Jesse-
