Kubernetes¶

Plain manifests, applied with Kustomize (kubectl apply -k): a Deployment of one replica, its Service, a volume for the state store and the configuration as a ConfigMap.

Install¶

1. Write your accounts, jobs and output in config.yaml (the README of each provider package has an example), and check it:

tapyr validate --config config.yaml

1. Create the namespace and the Secret tapyr with the environment variables the configuration refers to ({ env: NAME }) and TAPYR_STATE_KEY, which encrypts the tokens kept in the state store. From a file of NAME=value lines, kept out of Git:

kubectl create namespace tapyr
kubectl -n tapyr create secret generic tapyr --from-env-file=tapyr.env

1. Apply:

kubectl apply -k .
kubectl -n tapyr rollout status deploy/tapyr
kubectl -n tapyr port-forward svc/tapyr 9790:9790    # then http://127.0.0.1:9790/ui/

Operate¶

• Configuration changes. The name of the ConfigMap carries a hash of config.yaml: edit the file, apply again, and the Deployment restarts with the new configuration. A restart keeps the state store, so nothing is spent twice.
• One replica. Never scale the Deployment: a second instance would spend the same quotas without coordinating. The Recreate strategy stops the old pod before the new one opens the state store.
• State. Keep the volume: it remembers what was already spent of the daily and monthly quotas, and the rotated tokens (every refresh of an Aruba Central token returns a new refresh token, and the previous one is no longer valid).
• Resources. About 75 MiB of memory for 22,000 series (10,000 devices): adjust the limit to the size of your inventories.
• Logs. kubectl -n tapyr logs deploy/tapyr, in JSON.

Scrape¶

With the Prometheus Operator, a ServiceMonitor for the collected samples and the self-metrics:

apiVersion: monitoring.coreos.com/v1
kind: ServiceMonitor
metadata:
  name: tapyr
  namespace: tapyr
spec:
  selector:
    matchLabels: { app.kubernetes.io/name: tapyr }
  endpoints:
    - port: http
      path: /metrics
      honorTimestamps: true
    - port: http
      path: /metrics/self

To push instead of being scraped, add a remote_write or otlp output to config.yaml (see docs/outputs.md).