Replica cannot connect to core

[helanyao]

Please keep the following things in mind:

Hi admin,
I have installed neo4j 4.0.4 by helm chart

github.com

neo4j-contrib/neo4j-helm/blob/master/user-guide/USER-GUIDE.md

# Neo4j-Helm User Guide

## Overview

Neo4j-helm allows users to deploy multi-node Neo4j Enterprise Causal Clusters to Kubernetes instances, with configuration options for the most common scenarios.  It represents a very rapid way to get started running the world leading native graph database on top of Kubernetes.

This guide is intended only as a supplement to the [Neo4j Operations Manual](https://neo4j.com/docs/operations-manual/4.0/?ref=googlemarketplace).   Neo4j-helm is essentially a docker container based deploy of Neo4j Causal Cluster.  As such, all of the information in the Operations Manual applies to its operation, and this guide will focus only on kubernetes-specific concerns.

## Prerequisites

* Kubernetes 1.6+ with Beta APIs enabled
* Docker and kubectl installed locally
* Helm >= 3.1 installed
* PV provisioner support in the underlying infrastructure
* Requires the following variables
  You must add `acceptLicenseAgreement` in the values.yaml file and set it to `yes` or include `--set acceptLicenseAgreement=yes` in the command line of helm install to accept the license.
* This chart requires that you have a license for Neo4j Enterprise Edition.  Trial licenses 
[can be obtained here](https://neo4j.com/lp/enterprise-cloud/?utm_content=kubernetes)

## Licensing & Cost

This file has been truncated. show original

After the cluster is up, the replica cannot connect to the core.
root@c-neo4j-replica-cbb56c96f-f7lpv:/var/lib/neo4j/bin# ./cypher-shell
username: neo4j
password: ********
Failed to obtain connection towards WRITE server. Known routing table is: Ttl 1591175610069, currentTime 1591175310075, routers AddressSet=[10.42.0.1:7687], writers AddressSet=, readers AddressSet=[10.42.0.1:7687], database ''

Here is the cluster info

debug.log

[david_allen · Accepted solution]

use the bolt:// scheme to connect to an individual read replica. The replica does not have a full copy of the cluster's upstream routing table, and so it is true that when you use neo4j:// as a connection scheme, you cannot obtain a connection to a WRITE server.

More information:

https://neo4j.com/docs/driver-manual/current/client-applications/#driver-configuration-examples

[helanyao]

Thanks. You are right, the replica can be connected through bolt.