RouterBaba

// Nine labs, one curriculum

Break BGP on purpose. Understand it for life.

Each lab gives you a topology and a problem. You type real router commands. The protocol responds. You finish only when the network does what it should.

0 of 9 labs complete
Part I

Session establishment

What does it actually take for two routers to start talking?

01 BGP

First Hello

Vancouver and Calgary should be peering. They're not. Open the console, type one command, watch the FSM walk from Idle to Established.

  • 15 min
  • Beginner
  • Idle to Established
Enter the lab
02 BGP

The Mismatched Handshake

The session reaches OpenSent and dies. NOTIFICATION code 2 subcode 2. Read the protocol microscope. Find the ASN someone fat-fingered.

  • 20 min
  • Beginner
  • OPEN negotiation
Enter the lab
03 BGP

The Hold Timer Trap

The session establishes, then dies every three minutes. NOTIFICATION code 4. Two routers, two different timer configs, one negotiated minimum.

  • 25 min
  • Beginner
  • KEEPALIVE / Hold
Enter the lab
Part II

Path manipulation

You have multiple paths. Which one wins, and why?

04 BGP

Why The Long Path?

Two paths to Toronto. The shorter one should win. It doesn't. Local Preference runs before AS_PATH length. Find the route-map.

  • 25 min
  • Intermediate
  • LOCAL_PREF
Enter the lab
05 BGP

The MED Misunderstanding

Lower MED should win. It doesn't. Welcome to the most-misunderstood BGP attribute. IGP cost, always-compare-med, and why this trips up senior engineers.

  • 30 min
  • Intermediate
  • MED + IGP
Enter the lab
06 BGP

Communities and Policy

Customer routes are leaking to peers. They shouldn't. Write the route-maps that match communities and stop the leak. The internet runs on this.

  • 30 min
  • Intermediate
  • Communities
Enter the lab
Part III

Scale and operation

How do you run BGP without it eating your weekend?

07 BGP

The Route Reflector

Six iBGP routers. Full-mesh would be 15 sessions. Configure Toronto as a route reflector. Five sessions later, every router has every prefix.

  • 25 min
  • Advanced
  • iBGP scaling
Enter the lab
08 BGP

Convergence Under Failure

A link fails. Traffic drops for 90 seconds. Why didn't BGP notice? Fast-fallover, BFD, and the difference between protocol-time and human-time.

  • 25 min
  • Advanced
  • Fast convergence
Enter the lab
09 CAPSTONE

The Production Outage

3am. PagerDuty woke you up. Multiple things are wrong. Diagnose under pressure. Use everything you've learned. Write the post-mortem.

  • 45 min
  • Capstone
  • Integration
Enter the lab

// How a lab works

You. A topology. A real router. A problem.

  1. 01

    Open the lab

    Mission brief on the left, network map in the middle, router console on the right. The protocol is already running. Something is broken.

  2. 02

    Investigate

    Type real commands. show ip bgp summary. show ip bgp neighbor X.X.X.X. Click any packet on the wire to decode it. Click any concept to see why it works that way.

  3. 03

    Fix it

    Configure the missing piece. Watch the network respond in real time. The mission completes when the protocol does what it should, not when you click "done".

  4. 04

    Earn the next lab

    Each completion unlocks the next. Lab 09 unlocks only after all eight prerequisites. No shortcuts. No skip-ahead.

// Why this exists

Most networking courses teach you the answer. They don't teach you the question.

Reading about BGP is not the same as fixing a broken BGP session at 3am. These labs put you in front of a broken thing, hand you a console, and let the protocol grade your work. The protocol does not care about your feelings. That is the point.

Built by someone who has racked the gear and scaled the cloud. From hands-in-the-rack and pulling cable to hyperscaler-grade networking to advising national telcos. The labs are written from every layer of that stack, not from a textbook.