If you're setting up MySQL Cluster on Oracle Linux or another Linux such as Red Hat, CentOS or Fedora, you might have come across some problems getting the nodes to communicate. There are two ways you might bump into problems with network connectivity: The iptables firewall, and SELinux. These security mechanisms might prevent your various nodes—management, data, and API—from communicating with each other in various ways and with various symptoms.

Let's have a look at what you're likely to see. 

Data nodes stuck at "starting"

The first thing you might notice is that your data nodes get stuck in the "starting" state. Running show in the management client gives something like this:

[ndbd(NDB)]     2 node(s)
id=2    @192.168.56.211  (mysql-5.6.11 ndb-7.3.2, Nodegroup: 0, Master)
id=3    @192.168.56.212  (mysql-5.6.11 ndb-7.3.2, starting, Nodegroup: 0)

...and it never moves away from starting.

…