DevOps Coding Challenge #7: Parallel Pod Logger

Kubernetes Parallel Pod Logger: DevOps Challenge to Boost Observability Skills

Level: Intermediate
Languages: Python or Go
Focus: Kubernetes API, Concurrency, Observability

---

Problem Statement

You're debugging an issue affecting multiple pods across namespaces. You need to fetch logs for all matching pods in parallel to speed up troubleshooting.

Your task:
Write a CLI tool (in Python or Go) that:

• Accepts a label selector (e.g., app=backend)

• Runs across all namespaces

• Fetches logs from all matching pods concurrently

• Stores each log into a separate file in a local logs/ folder

---

Core Requirements

• Accept the following CLI flags:

  • --selector (label selector like app=web)

  • --since (optional, logs from last X minutes)

  • --limit (optional, max number of pods to fetch logs for)

• Fetch logs only from containers in Running state

• Use concurrent workers to fetch logs (e.g., goroutines in Go, ThreadPool in Python)

• Save logs in files named:
  logs/<namespace>_<pod-name>.log

---

Bonus Features

• Add a --container flag to specify which container to target (if pod has multiple)

• Add a progress bar (like tqdm in Python or pb in Go)

• Add retry logic for transient API errors

• Add --dry-run to show which pods would be logged

---

Tech Suggestions

Python:

• Use kubernetes Python client

• Use concurrent.futures.ThreadPoolExecutor

• Use argparse for CLI

• Use datetime to filter logs using since_seconds

Docs: https://github.com/kubernetes-client/python

Go:

• Use client-go

• Use Cobra for CLI

• Use context.WithTimeout

• Use sync.WaitGroup or goroutines with channels

Docs: https://github.com/kubernetes/client-go

---

Evaluation Criteria

• Efficient concurrent log fetching

• Clean folder structure and CLI UX

• Error handling and retries

• Bonus points for: container selection, dry-run, progress feedback

---