How Train Booking Platforms Handle Tatkal, WL, and RAC

“Tatkal gets sold out in seconds…
so what decides who actually gets the seat?”

During Tatkal, the system handles millions of requests and locks seats using extreme-speed concurrency control so each seat is allocated only once.
When all seats are full, the platform automatically places users in a Waiting List and updates it dynamically based on cancellations and RAC movement.
Final status — Confirmed, RAC, or WL — is calculated only after network-wide data processing.
If you didn’t know this before, share this post — students love this stuff.

Tatkal isn’t random — it’s distributed computing at peak speed.

• Seat Locking: Redis Atomic Locks
• Traffic Distribution: Kafka / RabbitMQ / SQS
• WL / RAC Movement: Probability Engine using live data
• Ledger: Clustered PostgreSQL
• Traffic Control: Akamai / CloudFront CDN

Tatkal is a masterclass in high-load, high-fairness system design.
What do you want next — IRCTC Payments / Booking Portal Architecture / Aadhaar Authentication?