Partitioning of ndbcluster

[Peter Mueller]

Hello, I try to find the partition algorithm of the ndbcluster.

My setup:
CREATE TABLE `test` (
`f1` varchar(3) NOT NULL DEFAULT '000',
`f2` varchar(3) NOT NULL DEFAULT '000',
`f3` varchar(3) NOT NULL DEFAULT '000',
`f4` varchar(3) NOT NULL DEFAULT '000',
`f5` varchar(3) NOT NULL DEFAULT '000',
PRIMARY KEY (`f1`,`f2`,`f3`))
ENGINE=ndbcluster
PARTITION BY KEY (f1, f2)
PARTITIONS 8;

Asking mysql I'm getting information about the partition:
explain partitions select * from test2 where f1='cccc' and f2='aaaa';

+----+-------------+-------+------------+------+
| id | select_type | table | partitions | type | ...
+----+-------------+-------+------------+------+
| 1 | SIMPLE | test2 | p6 | ref | ...
+----+-------------+-------+------------+------+

I have gathered information about the calculation of the partition (md5, modulo of partitions, ...). But I donn't get the pices together.

Can someone show me the way how to calculate the partition?

f1='cccc' and f2='aaaa' => Partition p6
f1='????' and f2='????' => Partition p?