MYSQLD -> NDB_MGMD, connection problems (no replies)

I have a strang issue in my test envirorment, the configuration is as followed:

OS RHEL 6.3
MySQL-Cluster-client-gpl-7.2.10-1.el6.x86_64
MySQL-Cluster-server-gpl-7.2.10-1.el6.x86_64

The server use following bind addresses:
192.168.0.10 --> NDB data node
192.168.0.99 --> NDB mgmt node
192.168.0.15 --> mysqld

config.ini
##########

[NDBD DEFAULT]
NoOfReplicas=1
DataMemory=10G
IndexMemory=2G
DiskPageBufferMemory=1048M
SharedGlobalMemory=384M
MaxNoOfExecutionThreads=4
RedoBuffer=32M
FragmentLogFileSize=256M
NoOfFragmentLogFiles=6

[NDB_MGMD]
NodeId=1
HostName=192.168.0.99
DataDir=/var/lib/mysql-cluster

[NDBD]
NodeId=2
HostName=192.168.0.10
DataDir=/var/lib/mysql-cluster/data

[NDBD]
NodeId=3
HostName=192.168.0.11
DataDir=/var/lib/mysql-cluster/data

[MYSQLD]
NodeId=4
HostName=192.168.0.15

[MYSQLD]
NodeId=5
HostName=192.168.0.16

my.cnf
######
[MYSQLD]
bind-address=192.168.0.15
ndbcluster
ndb-nodeid=4
ndb-connectstring=192.168.0.99
#default-storage-engine=NDB
ndb_autoincrement_prefetch_sz=320

[MYSQL_CLUSTER]
ndb-connectstring=192.168.0.99


My starting procedure is:
1. ndb_mgmd --initial -f /etc/config.ini
2. ndbmtd --ndb-mgmd-host=192.168.0.99 --initial --bind-address=192.168.0.10

So far so good:
ndb_mgm> show
Cluster Configuration
---------------------
[ndbd(NDB)] 2 node(s)
id=2 @192.168.0.10 (mysql-5.5.29 ndb-7.2.10, Nodegroup: 0, Master)
id=3 (not connected, accepting connect from 192.168.0.11)

[ndb_mgmd(MGM)] 1 node(s)
id=1 @192.168.0.99 (mysql-5.5.29 ndb-7.2.10)

[mysqld(API)] 2 node(s)
id=4 (not connected, accepting connect from 192.168.0.15)
id=5 (not connected, accepting connect from 192.168.0.16)

3. service mysql start

Now the strange thing happens, in my.cnf I use the bind-address=192.168.0.15 but it looks like before this binding happens the mysqld want to connect to the ndb_mgmd and this request will be rejected:

130212 12:22:52 mysqld_safe Starting mysqld daemon with databases from /var/lib/mysql
130212 12:22:52 [Note] Plugin 'FEDERATED' is disabled.
Configuration error: Error: Could not alloc node id at 192.168.0.99 port 1186: Connection with id 4 done from wrong host ip 192.168.0.99, exp ected 192.168.0.15(192.168.0.15).
130212 12:22:52 [ERROR] NDB: error (1) Configuration error: Error: Could not alloc node id at 192.168.0.99 port 1186: Connection with id 4 do ne from wrong host ip 192.168.0.99, expected 192.168.0.15(192.168.0.15).
130212 12:22:52 [ERROR] Plugin 'ndbcluster' init function returned error.
130212 12:22:52 [ERROR] Plugin 'ndbcluster' registration as a STORAGE ENGINE failed.
130212 12:22:52 InnoDB: The InnoDB memory heap is disabled
130212 12:22:52 InnoDB: Mutexes and rw_locks use GCC atomic builtins
130212 12:22:52 InnoDB: Compressed tables use zlib 1.2.3
130212 12:22:52 InnoDB: Using Linux native AIO
130212 12:22:52 InnoDB: Initializing buffer pool, size = 128.0M
130212 12:22:52 InnoDB: Completed initialization of buffer pool
130212 12:22:52 InnoDB: highest supported file format is Barracuda.
130212 12:22:52 InnoDB: Waiting for the background threads to start
130212 12:22:53 InnoDB: 1.1.8 started; log sequence number 1595675
130212 12:22:53 [Note] Server hostname (bind-address): '192.168.0.15'; port: 3306
130212 12:22:53 [Note] - '192.168.0.15' resolves to '192.168.0.15';
130212 12:22:53 [Note] Server socket created on IP: '192.168.0.15'.
130212 12:22:53 [Note] /usr/sbin/mysqld: ready for connections.

Has anybody an idea whats wrong here? I have tested the same thing with fedora 17 (same config files and MySQL packages) and everthing worked fine.