Skip to content

Latest commit

 

History

History
249 lines (209 loc) · 6.59 KB

README.md

File metadata and controls

249 lines (209 loc) · 6.59 KB
Raw

Introduction

We'll have project automation, actions and general scripts in this location.

Connect to backend API - http://redsol.eastus.cloudapp.azure.com/backend

Building docker demo app

az acr build \
             -t redsolacr.azurecr.io/demo:latest \
             -r redsolacr \
             -f ./config/docker/demo.dockerfile \
             --build-arg DOCKER_REGISTRY=redsolacr.azurecr.io \
             .

Depoly demo app

kubectl delete namespace demo-app && \
  kubectl apply -R -f ./config/k8s/demo

Building RedSol Backend and Db

./script/buildimages.sh

Deploy RedSol Backend and Db

  1. Update version in ./config/k8s/hc/Chart.yaml, value for appVersion.
  2. Run Script 
    ./script/deploy.sh

Notes

These are just some general notes that will work with the following alias;

alias k='kubectl --namespace human-connection-dev'

Login to k8s cluster

Clone the project-config repo and cd ./project-config.

source ./.env.dev
./script/login.sh

Connecting to the ports for development

The username and password can be access from AKS secretes however, you can also access them directly from the running app when you login with kubectl credentials.

Username:

export NEO4J_USERNAME=$(k exec $(k get pods --selector app=backend -o json |jq -r '.items[].metadata.name'| head -1) -- env | grep NEO4J_USERNAME=)

Password:

export NEO4J_PASSWORD=$(k exec $(k get pods --selector app=backend -o json |jq -r '.items[].metadata.name'| head -1) -- env | grep NEO4J_PASSWORD=)

Run these two commands in two seperate terminal windows.

Terminal 1:

kubectl --namespace human-connection-dev port-forward service/neo4j 7474:7474 7687:7687

Terminal 2:

kubectl --namespace human-connection-dev port-forward service/backend 4000:4000

These commands will remain running unitl you cancel them with ctrl-c. They are intended to open up local ports for you to access the graphql backend urls with localhost.

Connect over terminal

k exec --stdin --tty backend-7568bcd55-nzg94 -- /bin/bash

Installing troubleshooting tools on neo4j

apt-get update && \
apt-get install -y \
    vim \
    curl \
    procps \
    net-tools \
    wget \
    jq \
    htop \
    iputils-ping \
    netcat

Installing troubleshooting tools on backend

apk update && \
apk add \
    wget \
    vim \
    curl \
    procps \
    net-tools \
    wget \
    jq \
    htop \
    iputils \
    netcat-openbsd

Testing pod to pod communication from backend

nc neo4j.human-connection.svc.cluster.local 7474

type HELO in connection

Seed the database

This step should help with seeding the db for AKS cluster.

  1. Open a bash terminal and clone the JusticeInternational/Human-Connection repo or run /script/buildimages.sh. The buildimages.sh script will clone the repo to the .env/Human-Connection folder.

    git clone https://github.com/JusticeInternational/Human-Connection ./.env/Human-Connection

  2. Make sure your logged into the target AKS cluster:

    source ./.env.dev
./script/login.sh

    If you run the following commands you should see the running pods:

    ❯ kubectl --namespace human-connection get pods
NAME                      READY   STATUS    RESTARTS   AGE
backend-7568bcd55-d6zz2   1/1     Running   0          24h
neo4j-664544488-6lhdz     1/1     Running   0          24h

  3. Proxy the connection to AKS neo4j app:

    alias k='kubectl --namespace human-connection'
k port-forward service/neo4j 7474:7474 7687:7687

    You should see output as follows:

    Forwarding from 127.0.0.1:7474 -> 7474
Forwarding from [::1]:7474 -> 7474
Forwarding from 127.0.0.1:7687 -> 7687
Forwarding from [::1]:7687 -> 7687

  4. Open a new bash terminal and setup the env. Copy the .env.template file to .env:

    cd ./.env/Human-Connection/backend && \
cp .env.template .env

    Check the username and password configured in /config/k8s/hc/secrets/hc.yaml, for production we'll be setting this up in Az KeyVault.

    export NEO4J_USERNAME='<value from data.NEO4J_USERNAME>'
export NEO4J_PASSWORD='<value from data.NEO4J_PASSWORD>'

    sed -i -e 's/NEO4J_USERNAME=.*/'${NEO4J_USERNAME}'/' ./.env && \
sed -i -e 's/NEO4J_PASSWORD=.*/'${NEO4J_PASSWORD}'/' ./.env

  5. You should have yarn and node setup for the next step. Run the command:

    cd ./backend
yarn install

    yarn db:seed

    You should see output as follows:

    ❯ yarn db:seed
yarn run v1.22.10
$ babel-node src/db/seed.js
Warning: Email middleware will not try to send mails.
Warning: Sentry middleware inactive.
Seeded Data...
✨  Done in 52.35s.

  6. Seed the RedSol data:

    > yarn db:seed:redsol
yarn run v1.22.15
$ babel-node src/db/seed-redsol.js
Warning: Email middleware will not try to send mails.
Warning: Sentry middleware inactive.
Seeded Data...
Done in 4.62s.

  7. You should now be able to browse the db at http://localhost:7474/browser/

Test the API

We can test the API with bash and curl against the public dev url: http://redsol.eastus.cloudapp.azure.com/backend/. Note this step will not require kubectl port-forward commands or an az login.

  1. Open a bash terminal session and run the following scripts to give some helper functions: 
    function graphql() {
 query="${1}"
 query="$(echo $query|tr -d '\n')"
 curl -X POST \
   -s -H "Content-Type: application/json" \
   --data "{ \"query\": \"${query}\"}" \
   ${CLIENT_API}/graphql
}
function graphql_schema() {
 graphql "query IntrospectionQuery {  __schema { types { name } } }" |jq -r
}
  2. Setup CLIENT_API with export CLIENT_API=http://redsol.eastus.cloudapp.azure.com/backend
  3. Try some commands:
    • Query the schema with introspection, this requires DEBUG=true or NODE_ENV=true: 
      graphql_schema
    • Query for admin users 
      graphql "{User(role:admin){name}}" | jq -r '.data.User[]'
    • Query for categories 
      graphql "{Category{id,name}}"| jq -r '.data'
    • Query for service category 
      graphql "{ServiceCategory{id,name}}"| jq -r '.data'
    • Query for service 
      graphql "{Service{id,name}}"| jq -r '.data'