Jobnik: A Kubernetes-Native Job Triggering Service

Jobnik is a lightweight web service that enables triggering, listing, monitoring, and retrieving logs from Kubernetes Jobs via a REST API. Ideal for event-driven workflows, CI/CD pipelines, and custom job execution needs.

---

---

🚀 Features

• ✅ Trigger Kubernetes jobs with custom environment variables and arguments
• 🆕 Jobnik UI. TBD
• 🆕 List jobs with pagination and metadata
• 🆕 Fetch logs from job pods
• 🔁 Automatically generates unique job names
• 🧼 Cleans up completed jobs 30 seconds after success

---

🧱 Prerequisites

• ✅ A Kubernetes cluster with appropriate RBAC permissions to create, list, and delete jobs
• ✅ A base Job template deployed in your cluster
• ✅ Docker image for your job logic
• ✅ Optionally, use ArgoCD for managing Jobnik deployments

---

📦 Installation (Helm)

helm install jobnik ./helm/jobnik

Default Configuration

• 1 replica
• RBAC enabled
• Resource requests: 128Mi memory, 64m CPU
• Exposed as a ClusterIP service on port 80 targeting port 8080

---

📚 API Endpoints

1. Trigger a Job

POST /api/job

Creates and runs a new Kubernetes job based on an existing template.

Request Body

Field	Type	Required	Description
jobName	string	✅ Yes	Base Job name to clone and trigger
namespace	string	✅ Yes	Namespace where job will run
envVars	object (string)	❌ No	Key-value pairs of environment variables
args	array of strings	❌ No	Command-line arguments to pass to job

Example Request (cURL)

curl -X POST http://localhost:8080/api/job   -H "Content-Type: application/json"   -d '{
    "jobName": "test-job",
    "namespace": "default",
    "envVars": {
      "LOG_LEVEL": "debug",
      "TIMEOUT": "30s"
    },
    "args": ["--export", "--dry-run"]
  }'

Response

{
  "message": "Job test-job-run-1712345678-1234 triggered successfully",
  "jobName": "test-job-run-1712345678-1234",
  "namespace": "default"
}

---

2. List Jobs

GET /api/jobs

Fetches jobs with optional pagination and namespace filtering.

Query Parameters

Field	Type	Required	Description
namespace	string	❌ No	Namespace to list jobs from (or all)
limit	int	❌ No	Max number of jobs to return (default: 10)
offset	int	❌ No	Pagination offset (default: 0)

Example Request

curl "http://localhost:8080/api/jobs?namespace=default&limit=5&offset=0"

Response Headers

X-Total-Count: 25
X-Limit: 5
X-Offset: 0

JSON Response

{
  "total": 25,
  "limit": 5,
  "offset": 0,
  "count": 5,
  "jobs": [
    {
      "name": "test-job-run-1712345678-0001",
      "status": "succeeded"
    }
  ]
}

---

3. Get Job Logs

GET /api/job/logs

Fetch logs from a pod belonging to a Kubernetes job.

Query Parameters

Field	Type	Required	Description
jobName	string	✅ Yes	Name of the job
namespace	string	❌ No	Namespace (default: default)
container	string	❌ No	Container name (if job has multiple)

Example Request

curl "http://localhost:8080/api/job/logs?jobName=test-job-run-1712345678&namespace=default"

JSON Response

{
  "jobName": "test-job-run-1712345678",
  "namespace": "default",
  "logs": "Starting job...
Step 1 done...
Job completed."
}

---

🧪 Sample Usage

🔧 Inside Kubernetes (Internal Cluster Access)

Replace job-service.default.svc.cluster.local with your actual service name.

✅ Python

import requests

url = "http://job-service.default.svc.cluster.local:8080/api/job"
payload = {
  "jobName": "test-job",
  "namespace": "default",
  "envVars": {
    "LOG_LEVEL": "debug"
  },
  "args": ["--export"]
}
headers = {"Content-Type": "application/json"}
response = requests.post(url, json=payload, headers=headers)
print(response.json())

---

✅ Go

package main

import (
  "bytes"
  "encoding/json"
  "fmt"
  "net/http"
)

func main() {
  data := map[string]interface{}{
    "jobName":   "test-job",
    "namespace": "default",
    "envVars": map[string]string{
      "LOG_LEVEL": "debug",
    },
    "args": []string{"--dry-run"},
  }

  jsonData, _ := json.Marshal(data)
  req, _ := http.NewRequest("POST", "http://job-service.default.svc.cluster.local:8080/api/job", bytes.NewBuffer(jsonData))
  req.Header.Set("Content-Type", "application/json")

  client := &http.Client{}
  resp, _ := client.Do(req)
  fmt.Println("Status:", resp.StatusCode)
}

---

✅ cURL

curl -X POST "http://job-service.default.svc.cluster.local:8080/api/job"      -H "Content-Type: application/json"      -d '{
        "jobName": "test-job",
        "namespace": "default",
        "envVars": {
            "LOG_LEVEL": "debug"
        },
        "args": ["--dry-run"]
     }'

---

📌 Notes

• Make sure the base job you're cloning is pre-created in the target namespace.
• Jobs are monitored in the background and deleted 30 seconds after successful completion.
• You can extend Jobnik to support more features like retries, notifications, or parallel jobs.