1

I have 2 hosts on which I am running Docker, I have successfully started a Geode Locator and Server on my first host.

However, when I try to start a new Geode locator on the 2nd server:

start server --name=aloha --locators=myCache[10334]

I get the following error:

[info 2017/11/18 10:44:17.228 UTC aloha <main> tid=0x1] Starting membership services

[info 2017/11/18 10:44:17.304 UTC aloha <main> tid=0x1] JGroups channel created (took 76ms)

[info 2017/11/18 10:44:17.317 UTC aloha <main> tid=0x1] GemFire P2P Listener started on /172.17.0.1:41936

[info 2017/11/18 10:44:17.318 UTC aloha <Geode Failure Detection Server thread 0> tid=0x1e] Started failure detection server thread on /172.17.0.1:55542.

[info 2017/11/18 10:44:17.334 UTC aloha <main> tid=0x1] Attempting to join the distributed system through coordinator 172.17.0.7(myCache:77)<ec><v0>:1024 using address 172.17.0.1(aloha:664):1024

[info 2017/11/18 10:44:29.346 UTC aloha <main> tid=0x1] Probable coordinator is still 172.17.0.7(myCache:77)<ec><v0>:1024 - waiting for a join-response

[info 2017/11/18 10:44:41.351 UTC aloha <main> tid=0x1] Probable coordinator is still 172.17.0.7(myCache:77)<ec><v0>:1024 - waiting for a join-response

[info 2017/11/18 10:44:53.355 UTC aloha <main> tid=0x1] Probable coordinator is still 172.17.0.7(myCache:77)<ec><v0>:1024 - waiting for a join-response

[info 2017/11/18 10:45:05.359 UTC aloha <main> tid=0x1] Probable coordinator is still 172.17.0.7(myCache:77)<ec><v0>:1024 - waiting for a join-response

[info 2017/11/18 10:45:17.361 UTC aloha <main> tid=0x1] Stopping membership services

[info 2017/11/18 10:45:17.362 UTC aloha <main> tid=0x1] GMSHealthMonitor server socket is closed in stopServices().

[info 2017/11/18 10:45:17.362 UTC aloha <Geode Failure Detection Server thread 0> tid=0x1e] GMSHealthMonitor server thread exiting

[info 2017/11/18 10:45:17.363 UTC aloha <main> tid=0x1] GMSHealthMonitor serverSocketExecutor is terminated

I am exposing the following ports: - 7575, 1099, 10334, 8045 and 1024

I have also added a firewall rule to allow traffic (tcp/udp) through 1024

But it seems that the aloha server that I wish to start it is waiting for the myCache response which never arrives.

Has anyone experienced this? Or does anyone have a solution for this?

Sergiu
  • 2,928
  • 3
  • 27
  • 37
  • I might have misunderstood the situation here, but it looks like the server is looking for the myCache locator inside the docker network while (and here comes the misunderstanding) it's actually running on a remote server. – Stefano Nov 18 '17 at 11:40
  • @Stefano I guess that might be it because otherwise I cannot explain it, I have tried to use names or the host IP, however, I wasn't able to do so for the locator as I am always getting this IP doesn't exist on this host – Sergiu Nov 19 '17 at 08:18
  • Is there a way of allowing the locator to use a hostname or the main host IP address instead of the container's IP address? (even 0.0.0.0) isn't possible – Sergiu Nov 19 '17 at 08:20
  • I would suggest trying to use a docker swarm with an overlay network which would ease a *lot* of networking issues you might run into. A typical Geode cluster would run within a single open network. By running on docker across hosts you're effectively having to route separate networks and deal with port issues. Using swarm you'll get a flat network across hosts and all of the containers will be able to communicate freely. – Jens D Nov 20 '17 at 15:23
  • @JensD Thanks a lot for the comment. Yes that would make more sense as it would be joined in a network which would also give me the possibility to scale the service as required and moreover make my life easier as you said not having to deal with port issues. – Sergiu Nov 20 '17 at 17:01
  • 1
    @GhostCat Sure, thanks for the heads up :) – Sergiu Oct 26 '18 at 15:23
  • I appreciate the quick and kind comeback! – GhostCat Oct 26 '18 at 15:44

0 Answers0