Checklist of multicast settings

If you believe your caching is not working in your cluster, go through the following checklist to help debug the problem.

  1. Issue the ping 224.0.0.1 command to see if the machines on your network are listening to multicast traffic. You can run ping 224.0.0.1 | grep <network IP of other server> for example, ping 224.0.0.1 | grep 192.168.1.100 to see if the given server in the cluster is listening to your server.
  2. Modify your network adapter of your local Linux® machine to allow multicast traffic. Here is an example when submitting ifconfig. In this example multicast is configured:
    en5: flags=5e080863,c0<UP,BROADCAST,NOTRAILERS,RUNNING,SIMPLEX,MULTICAST,GROUPRT,64BIT,CHECKSUM_OFFLOAD(ACTIVE),PSEG,LARGESEND,CHAIN>
    inet 9.156.178.105 netmask 0xffffff00 broadcast 9.156.178.255
       tcp_sendspace 131072 tcp_recvspace 65536 rfc1323 0
    lo0: flags=e08084b<UP,BROADCAST,LOOPBACK,RUNNING,SIMPLEX,MULTICAST,GROUPRT,64BIT>
    inet 127.0.0.1 netmask 0xff000000 broadcast 127.255.255.255
       inet6 ::1/0
  3. On some machines you may have to add a route for multicast traffic. This applies to AIX® machines as well as some Linux distributions. Here is an example of how this could look:
    Note: 10.0.0.10 is my default gateway. This is optional and the command may look different on your machine.
    route add -net 224.0.0.0/3 10.0.0.10
  4. Check that you can run MulticastHelloWorld successfully to verify there is no security or configuration issues at the network or server level.
  5. Make sure the configuration of the multicast_addr is the same for all of the machines in the cluster. It is easy to change one, and not the others. This can be verified in the native_cache.log file.
  6. The multicast address being used for the caching mechanism responsible to communicate such changes is being defined in the two configuration files: the $TOP/bin/conf/env_settings.ini file and after running the $TOP/bin/configure_env.sh command, also in the $TOP/etc/default/mdm-ehcache-config.xml file. Usually, you only have to modify your configuration $TOP/bin/conf/env_settings.ini file and invoke the $TOP/bin/configure_env.sh command. Usually, there is no need to configure the $TOP/etc/default/mdm-ehcache-config.xml file manually. The default address is 239.0.10.1 and the multicastGroupPort 4446. Speak with you network administrator if this address and port is free to use in your network and that there are no firewall or router rules that prevent multicast traffic.
  7. Try using a different multicast address, for example: 224.0.0.1.
  8. Ensure the ttl setting is set to 1 on all of the machines if using a cluster.
  9. Make sure that if you ping your machine name from that machine, that you get a valid IP address on the network and not 127.0.1.1.