Loving Couchbase, but run into a bit of an issue with Cross Datacenter Replication over a non local network.
I am running Couchbase 4.5.1 Community Edition.
I’ve been able to succesfully make XDCR work locally on the same network - that is two machines on the same physical network.
However, it fails to work when I try replicate with a remote Couchbase Server on another computer on a different physical network.
On the network for my remote computer, I have opened all the ports as described here.
I have also turned OFF the firewall for the remote computer as well.
I have also, likewise, on my computer port forwarded the same ports, as well as disabled my firewall.
I can remotely log into the Couchbase web console from both computers, as well as create the remote clusters without any issue.
Here they are setup - again, both consoles can be accessed by both computers, as well as remote clusters can be created.
I have tried protocol 1 and protocol 2.
The exact error message that they both create are:
192.168.100.190:capi_ce51f33355eb9c8deee98df8346bd748/doorkeeper/doorkeeper_10.0.0.130:8092_0:dial tcp 10.0.0.130:8092: i/o timeout
127.0.0.1:capi_ab57eb050d844ede479387bafd28bc21/doorkeeper/doorkeeper_192.168.100.190:8092_0:dial tcp 192.168.100.190:8092: i/o timeout
The only meaningful data I can get out of that - is that its trying a TCP connection and its timing out.
It also looks like it’s possibly trying to connect to the nodes IP, rather than the remote cluster.
PC 1 should be talking to 149...64, but the error message suggests its talking to 10.0.0.130, which is the local IP of that computer on the netowork.
Likewise, PC 2 should be talking to 110...184, whereas the logs say its trying to talk to 192.168.100.190, which is the local IP of that computer on its own network.
Hope I’ve provided enough information to get some meaningful help.
Thanks in advanced!