unit3

unit 3: update and upgrade plans

This unit builds on unit 2 showcasing how update, and upgrade plans get triggered. It also showcases how controlled parameter update can be triggered.

params.yaml

The sample has an additional TEST parameter, to showcase controlled parameter update. The trigger key of the TEST parameter points to the super plan to be executed on update.

NODE_COUNT:
  default: "3"
TEST:
  default: "0"
  trigger: super

templates

No changes here.

operator.yaml

tasks

Besides the deploy-task there is now also an update-task, upgrade-task, and super-task. At this point they all apply the same templates. This will change on further iteration, for example the upgrade-task will take care of the elasticsearch rolling upgrade steps.

deploy-task:
  resources:
    - service.yaml
    - node.yaml
update-task:
  resources:
    - service.yaml
    - node.yaml
upgrade-task:
  resources:
    - service.yaml
    - node.yaml
super-task:
  resources:
    - service.yaml
    - node.yaml

plans

The sample has not only the deploy plan, but also update, upgrade, and a super plans.

plans:
  deploy:
    strategy: serial
    phases:
      - name: deploy-phase
        strategy: parallel
        steps:
          - name: deploy-step
            tasks:
              - deploy-task
  update:
    strategy: serial
    phases:
      - name: update-phase
        strategy: parallel
        steps:
          - name: update-step
            tasks:
              - update-task
  upgrade:
    strategy: serial
    phases:
      - name: upgrade-phase
        strategy: parallel
        steps:
          - name: upgrade-step
            tasks:
              - upgrade-task
  super:
    strategy: serial
    phases:
      - name: super-phase
        strategy: parallel
        steps:
          - name: super-step
            tasks:
              - super-task

Run the framework instance

If you haven't already then clone the kudo-tutorial repository.

git clone https://github.com/realmbgl/kudo-tutorial.git

From the unit3/operator folder use the following command to run the instance.

kubectl kudo install . --instance myes

Once the install is finished we should see the following pods.

kubectl get pods
NAME          READY   STATUS    RESTARTS   AGE
myes-node-0   1/1     Running   0          42m
myes-node-1   1/1     Running   0          41m
myes-node-2   1/1     Running   0          41m

Lets check on the plan execution history using the kudo cli, we see the deploy plan has been executed.

kubectl kudo plan history --instance=myes
History of all plan-executions for instance "myes" in namespace "default":
.
└── myes-deploy-206753060 (created 4h18m37s ago)

Update the framework instance to scale to 4 nodes

Update the NODE_COUNT parameter to 4.

From the unit3/operator folder use the following command to update the instance.

kubectl kudo update --instance myes -p NODE_COUNT=4

Once the update is finished we should see an additional pod myes-node-3.

kubectl get pods
NAME          READY   STATUS    RESTARTS   AGE
myes-node-0   1/1     Running   0          46m
myes-node-1   1/1     Running   0          46m
myes-node-2   1/1     Running   0          45m
myes-node-3   1/1     Running   0          70s

Lets check on the plan execution history using the kudo cli, we see the update plan has been executed.

kubectl kudo plan history --instance=myes
History of all plan-executions for instance "myes" in namespace "default":
.
├── myes-deploy-206753060 (created 4h20m36s ago)
└── myes-update-207626186 (created 39s ago)

Upgrade the framework instance to use a newer framework implementation

Go to the unit3/operator-next folder. Its a copy of unit3/operator but with the operator version set to 0.2.0 in operator.yaml and using elasticsearch:7.2.0 for the image in node.yaml.

From the unit3/operator-next folder use the following command to upgrade the operator version.

kubectl kudo upgrade . --instance myes

operatorversion.kudo.dev/v1alpha1/elastic-0.2.0 successfully created
instance./myes successfully updated

Check on the versions available.

kubectl get operatorversion
NAME            AGE
elastic-0.1.0   33m
elastic-0.2.0   4s

Lets check on the plan execution history using the kudo cli, we see the upgrade plan has been executed.

kubectl kudo plan history --instance=myes
History of all plan-executions for instance "myes" in namespace "default":
.
├── myes-deploy-206753060 (created 4h22m0s ago)
├── myes-update-207626186 (created 2m3s ago)
└── myes-upgrade-521226614 (created 17s ago)

Lets see whether the myes pods use the newer elasticsearch container image, you should see that version 7.2.0 is used after the upgrade.

kubectl get pod myes-node-0 -o yaml | grep "docker.io/library/elasticsearch:"
    image: docker.io/library/elasticsearch:7.2.0

Controlled parameter update

Update the TEST parameter to like 100.

Use the following command to update the instance.

kubectl kudo update myes -p TEST=100

Lets check on the plan execution history using the kudo cli, we see the super plan has been executed.

kubectl kudo plan history --instance=myes
History of all plan-executions for instance "myes" in namespace "default":
.
├── myes-deploy-206753060 (created 4h23m46s ago)
├── myes-super-133145198 (created 4s ago)
├── myes-update-207626186 (created 3m49s ago)
└── myes-upgrade-521226614 (created 2m3s ago)