I am currently trying to build a table that will hold several billion rows small rows of data. From the info mysql cluster seems to be the way to go. However I am doing some testing and inserts into my set up are painfully slow. I have 4 2.4G Xeon 1G boxes.
If I create my table using myism I can insert a million rows in about 1 min. However if I use the ndbcluster engine, the same script take more than 15min to insert 1 million rows. I am on a GIG network.
Any suggestions on how to optimize these inserts would be very helpful.