The Swiss Army Knife for Developing Cloud Native Applications - Rio Cloud Conference Day

A demo to show how sync local Quarkus sources with pod running in kubernetes and use the quarkus:dev mode to hot reload the changes, we're using Vagrant, Microk8s, Tilt and Quarkus. During the environment provisioning we're starting Postgresql, Prometheus and Grafana (all of them are helm packages). It was built for a presentation done in Rio Cloud Conference Day with topic The Swiss Army Knife for Developing Cloud Native Applications. Slides in pt-br

Here is a summary of what is being done here (note that we have health checks properly setup). That's why first build takes more time.

Make a change at the code and see the magic happening:

Authors

• @bandrade
• @msmagnanijr

Important

This Dockerfile is used for development purposes, in order to have a production image, you may consider to use Native builds

Pre-requisite

• Have Docker or Buildah properly installed.
• Have Vagrant installed with vagrant-ssh plugin.
• Have Virtual box or Libvirt installed to create VMs.

Running the demo

Clone the demo sources locally:

https://github.com/bandrade/rio-cloud-conference-day.git
cd rio-cloud-conference-day/infra

Start the provisioning with the following command:

vagrant up

Waiting until the provisioning is done:

==> default: Setting hostname...
==> default: Automatic installation for Landrush IP not enabled
==> default: Installing SSHFS client...
==> default: Mounting SSHFS shared folder...
==> default: Mounting folder via SSHFS: /home/bandrade/Downloads/repo/infra => /home/vagrant/scripts
==> default: Checking Mount..
==> default: Checking Mount..
==> default: Folder Successfully Mounted!
==> default: Running provisioner: ansible...
Vagrant has automatically selected the compatibility mode '2.0'
according to the Ansible version installed (2.9.9).

Alternatively, the compatibility mode can be specified in your Vagrantfile:
https://www.vagrantup.com/docs/provisioning/ansible_common.html#compatibility_mode

    default: Running ansible-playbook...

PLAY [all] *********************************************************************

TASK [Gathering Facts] *********************************************************
ok: [default]

TASK [microk8s : check if this is really ubuntu] *******************************
skipping: [default]

TASK [microk8s : setting noninteractive mode] **********************************
changed: [default]

TASK [microk8s : install snap] *************************************************
ok: [default]

TASK [microk8s : check is ufw installed] ***************************************
changed: [default]

TASK [microk8s : disabling ufw] ************************************************
changed: [default]

TASK [microk8s : install microk8s] *********************************************
changed: [default]

TASK [microk8s : microk8s status] **********************************************
changed: [default]

TASK [microk8s : enable dns] ***************************************************
changed: [default]

TASK [microk8s : enable ingress] ***********************************************
changed: [default]

TASK [microk8s : enable registry] **********************************************
changed: [default]

TASK [microk8s : enable storage] ***********************************************
changed: [default]

TASK [microk8s : create kube dir] **********************************************
ok: [default]

TASK [microk8s : configure kubeconfig] *****************************************
changed: [default]

TASK [microk8s : configure snap alias] *****************************************
changed: [default]

TASK [microk8s : configure vagrant user] ***************************************
changed: [default]

PLAY RECAP *********************************************************************
default                    : ok=15   changed=12   unreachable=0    failed=0    skipped=1    rescued=0    ignored=0 

Export the KUBECONFIG environment variable and ensure that all pods are in Running state:

export KUBECONFIG=$(pwd)/.kube/kubeconfig
kubectl get pods --all-namespaces
NAMESPACE            NAME                                      READY   STATUS    RESTARTS   AGE
container-registry   registry-7cf58dcdcc-g8fc2                 1/1     Running   0          7m56s
ingress              nginx-ingress-microk8s-controller-kp2pm   1/1     Running   0          7m47s
kube-system          coredns-588fd544bf-s4j5n                  1/1     Running   0          8m6s
kube-system          hostpath-provisioner-75fdc8fccd-dcdc6     1/1     Running   0          7m55s

Now go to the app directory:

cd ../app

If it's being used buildah as builder, start the environment with:

tilt up

If it's being used docker as builder, start the environment with:

tilt up -f Tiltfile.docker

The command creates the following resources:

• a prometheus deployment
• a grafana deployment with some dashboards being imported
• a postgresql deployment
• a kubernetes deployment called quarkus-todo-demo
• a persistence volume claim called m2-cache to cache the maven artifacts for faster builds

Check the stats of the application:

kubectl get pods -w ##(1)

1. A successful start will have quarkus-todo-demo in Running status

Note	You can terminate the kubectl get pods -w using the command CTRL+C

You can now make changes to the sources e.g change Sort by to title in TodoResource.java

After few seconds you will see your TODO itens being ordered by name.

Cleanup

tilt down -f <Tiltfile>