Can't connect to nats streaming server in Docker cluster

It seems I get a connection time out every time I try to connect to a cluster on the docker container. I have no problem connecting with the Nats Client. The streaming client works until I turn on clustering.

Here’s how I fire up the cluster.

docker network create --driver bridge nats-cluster
docker run --network=nats-cluster -it --name a -p 0.0.0.0:4222:4222 -p 0.0.0.0:4248:4248 -v C:/data:/data nats-streaming -cid mycluster -st FILE -dir /data/a -cluster_log_path /data/mycluster --clustered --cluster_node_id a --cluster_peers b --cluster nats://0.0.0.0:4248 -routes nats-route://b:5248 -SDV

Here’s how I try to connect

var factory = new StanConnectionFactory();
var opts = StanOptions.GetDefaultOptions();
opts.NatsURL = "nats://localhost:4222";
var connection = factory.CreateConnection("mycluster", "mytestapp", opts);

Here’s the exception

STAN.Client.StanConnectRequestTimeoutException: ‘Connection Request Timed out.’

The non streaming client works fine. No errors doing this:

var fac = new ConnectionFactory();
var opts = ConnectionFactory.GetDefaultOptions();
opts.Url = "nats://localhost:4222";
fac.CreateConnection(opts);

navigating to localhost:8222 still brings up the monitor, I just can’t make a connection using the nats streaming client.

Am I doing something wrong?

[bennyr103]

Oops, turns out the cluster log path can’t be the same for all three containers when they’re running locally! I guess that’s why those commands were working fine from a windows command prompt but not from the docker containers.

The streaming client connects fine now that the cluster is properly configured.

If anyone sees this and just wants a quick simple local setup using docker containers here are the commands

docker network create --driver bridge nats-cluster

docker run --network nats-cluster --name nats-streaming-a -p 0.0.0.0:4222:4222 --rm -v C:/ContainerData:/ContainerData nats-streaming -p 4222 -cid nats-cluster --clustered --cluster_node_id nats-streaming-a --cluster_peers nats-streaming-b,nats-streaming-c -st FILE -dir /ContainerData/nats-streaming-a -cluster_log_path /ContainerData/nbs-cluster1 --cluster nats://0.0.0.0:4248 -routes nats://nats-streaming-b:5248,nats://nats-streaming-c:6248

docker run --network nats-cluster --name nats-streaming-b -p 0.0.0.0:5222:5222 --rm -v C:/ContainerData:/ContainerData nats-streaming -p 5222 -cid nats-cluster --clustered --cluster_node_id nats-streaming-b --cluster_peers nats-streaming-a,nats-streaming-c -st FILE -dir /ContainerData/nats-streaming-b -cluster_log_path /ContainerData/nbs-cluster2 --cluster nats://0.0.0.0:5248 -routes nats://nats-streaming-a:4248,nats://nats-streaming-c:6248

docker run --network nats-cluster --name nats-streaming-c -p 0.0.0.0:6222:6222 --rm -v C:/ContainerData:/ContainerData nats-streaming -p 6222 -cid nats-cluster --clustered --cluster_node_id nats-streaming-c --cluster_peers nats-streaming-b,nats-streaming-a -st FILE -dir /ContainerData/nats-streaming-c -cluster_log_path /ContainerData/nbs-cluster3 --cluster nats://0.0.0.0:6248 -routes nats://nats-streaming-b:5248,nats://nats-streaming-a:4248

Streaming client works like this:

var factory = new StanConnectionFactory();
var sopts = StanOptions.GetDefaultOptions();
sopts.NatsURL = "nats://localhost:4222,nats://localhost:5222,nats://localhost:6222";
var connection = factory.CreateConnection("nats-cluster", "mytestapp", sopts);

Thanks for the responses!

[bennyr103]

With the --cluster_raft_logging I’m seeing a lot of these

[1] 2019/03/04 23:34:31.689149 [TRC] STREAM: raft: Vote granted from nats-streaming-a in term 55. Tally: 1
[1] 2019/03/04 23:34:33.690077 [ERR] STREAM: raft: Failed to make RequestVote RPC to {Voter nats-streaming-c nbs-cluster.nats-streaming-c.nbs-cluster}: nats: timeout
[1] 2019/03/04 23:34:33.690353 [ERR] STREAM: raft: Failed to make RequestVote RPC to {Voter nats-streaming-b nbs-cluster.nats-streaming-b.nbs-cluster}: nats: timeout
[1] 2019/03/04 23:34:35.515287 [WRN] STREAM: raft: Election timeout reached, restarting election
[1] 2019/03/04 23:34:35.515354 [INF] STREAM: raft: Node at nbs-cluster.nats-streaming-a.nbs-cluster [Candidate] entering Candidate state in term 56
[1] 2019/03/04 23:34:35.515747 [TRC] STREAM: raft: Votes needed: 2

This probably isn’t a client issue anymore, is it…

Do my Docker run commands look correct?