Seeking help in getting the browser interface working

[Analect]

@mneedham

Thanks for your efforts in getting this running. I wanted to test it out, but I've hit a few obstacles.
I think your README is slightly out-of-sync with your files in this repo, but I was able to get things going by running kubectl create -f neo4j_env.yml, which combines the core cluster statefulset and a service together.

As per these logs from one of the core pods, the cluster seems to form fine.

2017-02-09T13:20:13.558079356Z Starting Neo4j.
2017-02-09T13:20:14.882575065Z 2017-02-09 13:20:14.881+0000 INFO  No SSL certificate found, generating a self-signed certificate..
2017-02-09T13:20:16.123709705Z 2017-02-09 13:20:16.123+0000 INFO  Starting...
2017-02-09T13:20:18.015336265Z 2017-02-09 13:20:18.014+0000 INFO  Bolt enabled on 0.0.0.0:7687.
2017-02-09T13:20:18.042945087Z 2017-02-09 13:20:18.042+0000 INFO  Initiating metrics...
2017-02-09T13:20:18.363854719Z 2017-02-09 13:20:18.363+0000 INFO  Waiting for other members to join cluster before continuing...
2017-02-09T13:21:16.087812697Z 2017-02-09 13:21:16.087+0000 INFO  Started.
2017-02-09T13:21:16.531841102Z 2017-02-09 13:21:16.531+0000 INFO  Mounted REST API at: /db/manage
2017-02-09T13:21:19.500402108Z 2017-02-09 13:21:19.500+0000 INFO  Remote interface available at http://neo4j-core-0.neo4j.default.svc.cluster.local:7474/

However, in looking to interface with the installation, I ran kubectl port-forward neo4j-core-0 7474:7474, which forwards port 7474 from the cluster to my local browser. While this appears to partly work, it's not allowing me to enter the default neo4j password, so that I can set my own.

In the past, I've been able to get around this problem by adding an end-point for bolt. I modified the headless service as follows:

# Headless service to provide DNS lookup
apiVersion: v1
kind: Service
metadata:
  labels:
    app: neo4j
  name: neo4j
spec:
  clusterIP: None
  ports:
    - port: 7474
      name: http
      targetPort: 7474
    - port: 7687
      name: bolt
      targetPort: 7687  
  selector:
    app: neo4j

... hoping that that would allow me to feed either 0.0.0.0:7687 or neo4j-core-0.neo4j.default.svc.cluster.local:7687 into the bolt field in settings, but neither work. Can you suggest what I might be able to do to get this working? Is there another form/way of entering the bolt end-point? I see you have value: "bolt://@192.168.1.3:7687" at the bottom of the README, but I'm not sure in what context.

In terms of getting pre-existing data into this cluster, is there a recommended approach using kubectl exec to stop a neo4j process temporarily on one node and then kubectl cp to copy in a backup/zipped folder of a database, that can be used to seed the other nodes?

[sherodtaylor]

@Analect did you get this working?

[Analect]

I think I just opted for the "don't use bolt" in the settings to get it working.

[lolstarz]

Also having the same problem and hoping there's a fix someone is familiar with

[ghost]

I am also facing the same issue. Can anybody help to fix the same and finally access using (value: "bolt://@192.168.1.3:7687") this url format

[mneedham]

@Analect can you try port forwarding the Bolt port as well?

kubectl port-forward neo4j-core-0 7474:7474 7687:7687

That should then allow the Neo4j browser to connect to the Bolt port on the Kubernetes pod

[Analect]

Thanks @mneedham ... haven't really touched the kubernetes solution for neo4j in a long time. Is this helm chart now considered the best approach. Never really found a good way of populating/seeding a pre-running neo4j in kubernetes ... all approaches required a stopped instance, which was hard to reconcile in kubernetes ... don't suppose there's a better solution to that now? Thanks.