Dist Cache in a clustered nodes

[mahesh.vedantam]

Hi,

I am not able to synch up cache in a dist cache in a clustered environment.

I have configured Node.java to use infinispan.xml and JGroups configuration using TCP protocol.

Node A is started from IP1 on machine1 in LAN --- port 7800

Node B is started from IP1 on machine2 in LAN --- port 7801

For some reason, the cache contents are NOT @replicated/copied once the data is updated/inserted into the cache

and I can see the contents of the cache on their respective machine on which its started.

However, I was able to synchup the contents by launching 3 processes on the same machine(binding to loopback).

JGroups.xml is configured to use the IP address of the machine and port(diff for each node), TCP, UNiCAST3, MPING,

and also using mcast ip address and mcast ip port. Its helpfulll if you can respond back.

regards, Mahesh.

[nadirx]

Ensure that you are binding to the correct IP address and that there are no iptables/firewall rules blocking traffic.

It would help if you posted your jgroups.xml and log files for  both machines.

[mahesh.vedantam]

I attached the jgroups.xml file( it is the same format for mach2), pls check.

jgroups.tcp.port: 7800(mach1) and 7801(mach2)

jgroups.bind_addr-<ip> and <ip2> for mach1 and mach2.

Please respond if any issues with xml files. I am using infinispan 9.0.0.Final version.

[mahesh.vedantam]

I was able to test out the TCP Server in java vs TCP Client on Mach1 and Mach2 (with different IP on LAN).

Message was transmitted successfully.

Whereas for the infinispan Node Process launched with IP1 and port1 and Ip2 & port2 is not working.

Is it possible to debug the processes ?

What sections of the jgroups needs to checked for Unicast TCP replication in dist cache? please advise.

Umamahesh.

[genman]

If UDP multicast isn't working you can set up TCP Ping:

http://www.jgroups.org/manual/html/user-advanced.html#d0e2494