K
Q

How can I access a nested item for a Helm charts custom value file?

4/8/2020

I'm using the stable/prometheus helm chart and I've configured a custom values file that further configures alertmanager for the chart deployment. I can install the chart via Helm3 without any issues, however, there's one thing I'm not able to figure out. For the Slack Reciever/slack_configs/api_url I want to pass that through the set command so I don't have to keep it hardcoded into file.

I hope I'm on the right path, and here's what I'm thinking of running to access the value.

helm install test-release stable/prometheus -f customALM.yml --set alertmanagerFiles.alertmanager.yml.receivers[0].api_url=https://hooks.slack.com/services/XXXXXXXXX/XXXXXXXXX/xxxxxxxxxxxxxxxxxxxxxxxxxxx

customALM.yml

alertmanagerFiles:
  alertmanager.yml:
        route:
          group_wait: 10s
          group_interval: 5m
          repeat_interval: 30m
          receiver: "slack"

          routes:
            - receiver: "slack"
              group_wait: 10s
              match_re:
                severity: error|warning
              continue: true

        receivers:
          - name: "slack"
            slack_configs:
              - api_url: '[howDoIsetThisAtTheCLI?'
                channel: 'someChannel'
                text: "Text message template etc."

Update 4/8: I'm making progress thanks to willro! I'm able to get a value inserted but it puts it at the root of the alertmanager block. I've tried a few different combinations to access receivers/slack_configs but no luck yet =/

helm install test-release stable/prometheus -f customALM.yml --set alertmanagerFiles.api_url=PleaseInsertPrettyPlease --dry-run

Update 4/9: I've decided to move the receivers block into a separate file that's encrypted and stored securely.

-- RomeNYRR
kubernetes
kubernetes-helm
yaml

Similar Questions

Error response from daemon: Get https://registry-1.docker.io/v2/: dial tcp: lookup registry-1.docker.io...i/o timeout
nginx ingress controller not reading configmap
Frequent Disconnection and re-connection of web socket requests after deployment in the K8s
Deploy EKS aws without downtime

1 Answer

4/10/2020

Running the command to change the URL after its been deployed is definitely an option that I want to have.

I'd like to write this answer to give you this option!

  • You can chain a few commands with SED to edit data on that ConfigMap (it's very similar to what kubectl edit does!)

  • For that you will need to use the string deployed on the customALM.yml. For this example I set the parameter as api_url: ChangeMeLater before deploying.

  • Then I deployed the chart with helm install test-release stable/prometheus -f customALM.yml

  • Lastly we run:

kubectl get cm <CONFIG_MAP_NAME> -o yaml | sed -e "s,<OLD_VALUE>,<NEW_VALUE>,g" | kubectl replace -f -

  • Explaining what's going on:

    • kubectl get cm <CONFIG_MAP_NAME> -o yaml | = gets the deployed configmap in yaml format and pipe it to the next command
    • sed -e "s,<OLD_VALUE>,<NEW_VALUE>,g" | = use sed to replace old_value for new_value and pipe it to the next comand
    • kubectl replace -f - = use the output from the last command and replace the object currently deployed with the same name.

  • I'll leave an example here step by step to elucidate more:

$ helm install test-release stable/prometheus -f customALM.yml
Release "test-release" has been installed. Happy Helming!
...

$ kubectl get cm
NAME                                   DATA   AGE
test-release-prometheus-alertmanager   1      44m
test-release-prometheus-server         5      44m

$ kubectl get cm test-release-prometheus-alertmanager -o yaml
apiVersion: v1
data:
  alertmanager.yml: |
    global: {}
    receivers:
    - name: slack
      slack_configs:
      - api_url: ChangeMeLater
        channel: someChannel
        text: Text message template etc.
    route:
      group_interval: 5m
      group_wait: 10s
      receiver: slack
      repeat_interval: 30m
      routes:
      - continue: true
        group_wait: 10s
        match_re:
          severity: error|warning
        receiver: slack
kind: ConfigMap
metadata:
  creationTimestamp: "2020-04-10T13:41:15Z"
  labels:
    app: prometheus
    chart: prometheus-11.0.6
    component: alertmanager
    heritage: Helm
    release: test-release
  name: test-release-prometheus-alertmanager
  namespace: default
  resourceVersion: "218148"
  selfLink: /api/v1/namespaces/default/configmaps/test-release-prometheus-alertmanager
  uid: 323fdd40-2f29-4cde-833c-c6300d5688c0

$ kubectl get cm test-release-prometheus-alertmanager -o yaml | sed -e "s,ChangeMeLater,theurl.com/any,g" | kubectl replace -f -
configmap/test-release-prometheus-alertmanager replaced

$ kubectl get cm test-release-prometheus-alertmanager -o yaml
apiVersion: v1
data:
  alertmanager.yml: |
    global: {}
    receivers:
    - name: slack
      slack_configs:
      - api_url: theurl.com/any
        channel: someChannel
        text: Text message template etc.
    route:
      group_interval: 5m
      group_wait: 10s
      receiver: slack
      repeat_interval: 30m
      routes:
      - continue: true
        group_wait: 10s
        match_re:
          severity: error|warning
        receiver: slack
kind: ConfigMap
metadata:
  creationTimestamp: "2020-04-10T13:41:15Z"
  labels:
    app: prometheus
    chart: prometheus-11.0.6
    component: alertmanager
    heritage: Helm
    release: test-release
  name: test-release-prometheus-alertmanager
  namespace: default
  resourceVersion: "219507"
  selfLink: /api/v1/namespaces/default/configmaps/test-release-prometheus-alertmanager
  uid: 323fdd40-2f29-4cde-833c-c6300d5688c0

You can see that the command changed the ChangeMeLater for theurl.com/any.

I'm still thinking about your first option, but this is a good workaround to have in hand.

If you have any doubt let me know!

-- willrof
Source: StackOverflow