Skip to content

Troubleshooting a Blocker

📝 Context

Something isn't working during implementation. You've hit a blocker and need to systematically troubleshoot it. Don't spin your wheels - follow a process and know when to ask for help.

📋 Before You Start Troubleshooting

  • [ ] Can you reproduce the issue consistently?
  • [ ] Do you have error messages or logs?
  • [ ] What were you trying to do when it failed?
  • [ ] What changed recently? (Your changes, environment changes, etc.)
  • [ ] Have you checked basic connectivity and access?

🎯 Troubleshooting Process

Step 1: Reproduce and Document

  • Can you reproduce it? If not, gather more information
  • What's the exact error message or behavior?
  • When does it happen? (Always, sometimes, specific conditions?)
  • What were you doing when it occurred?
  • Take screenshots, copy logs, note timestamps

Step 2: Isolate the Problem

  • Is it environment-specific? (Works elsewhere, fails here?)
  • Is it configuration-related? (Check your values, settings)
  • Is it network-related? (Can you reach required endpoints?)
  • Is it permission-related? (Do you have the right access?)
  • Is it version-related? (Kubernetes, Helm, dependencies?)

Step 3: Check the Basics

  • Environment access still working?
  • Network connectivity (can you reach what you need?)
  • Resource availability (CPU, memory, storage)
  • Service status (are required services running?)
  • Recent changes (what changed in the last hour/day?)

Step 4: Search and Research

  • Check documentation for known issues
  • Search internal knowledge base or past tickets
  • Check GitHub issues or release notes
  • Review environment-specific guides
  • Look for similar problems in team channels

Step 5: Timebox Your Debugging

  • Set a time limit ⏱️ 30-60 minutes
  • If you're not making progress, stop and ask for help
  • Don't spend hours on something others have solved
  • Document what you've tried before asking

🎯 Common Issue Categories

Deployment Issues

  • Pods not starting
  • Images not pulling
  • ConfigMaps/Secrets not mounting
  • Services not exposing correctly
  • Ingress not routing

Configuration Issues

  • Wrong values in Helm charts
  • Missing environment variables
  • Incorrect resource limits
  • Network policies blocking traffic
  • Storage classes not available

Access Issues

  • Can't access cluster
  • Permissions denied
  • Service account issues
  • RBAC misconfiguration
  • Network policies blocking

Integration Issues

  • Can't reach external services
  • Authentication failures
  • API version mismatches
  • Data format issues
  • Timeout problems

🎯 When to Ask for Help

Ask for help if:

  • You've spent 30-60 minutes without progress
  • You've tried the obvious fixes
  • The issue is blocking critical path
  • You don't have the expertise for this area
  • The customer is waiting

How to ask:

  • Use Requesting help framework
  • Provide context (what you're trying to do)
  • Share what you've tried
  • Include error messages and logs
  • Note urgency level

⚠️ Gotchas

  • Not reproducing first - you might be debugging the wrong thing
  • Not checking logs - error messages tell you a lot
  • Assuming it's your fault - could be environment, network, permissions
  • Spending too long - timebox and ask for help
  • Not documenting what you tried - wastes time when you ask for help
  • Panicking - most issues are solvable with a systematic approach

Built as a public field guide for practical Solutions Engineering and Architecture work.