Job monitor

This repo monitors finished or failed kubernetes jobs.

Project: https://github.com/dpetzold/kube-job-monitor

Docker image: https://registry.hub.docker.com/u/dpetzold/kube-job-monitor/

Usage

Command Line Options

• -failures - Threshold of allowable failures for a job:
  • -failures=0: The job will be reaped on any failures.
  • -failures=-1: The job will never be reaped on failures (default).
• -log - Level to log - debug, info, warn, error, fatal, panic (default info)
• -keep-completed=*duration* - Duration to keep completed jobs (e.g. -keep-completed=4h).
• -ignore-owned - Ignore jobs owned by other resources, e.g. by CronJobs (which have their own reaping logic).

Kubernetes Pod Definition

apiVersion: apps/v1
kind: Deployment
metadata:
  name: job-monitor
spec:
  replicas: 1
  selector:
    matchLabels:
      name: job-monitor
  template:
    metadata:
      labels:
        name: job-monitor
    spec:
      containers:
      - name: job-monitor
        image: dpetzold/job-monitor:latest

Development

The following will get you setup for local development:

mkdir -p $GOPATH/github/dpetzold
cd $GOPATH/github/dpetzold
git clone https://github.com/dpetzold/kube-job-monitor.git
minikube start
eval $(minikube docker-env)
make docker
kubectl apply -f examples/job-monitor.yaml