The user wants me to rewrite a headline about Airbnb's search and booking system avoiding double-booking.

Short summary

The article explains how lodging marketplaces like Airbnb prevent double-booking by treating reservations as distributed race conditions where search needs speed and eventual consistency while booking requires strong transactional consistency. It recommends modeling availability as date ranges with database-level exclusion constraints (Postgres GiST index) rather than application-level check-then-set, and separating the search index from the booking store. It also covers payment holds, idempotency keys, and the broader principle of not forcing reads and writes to share the same consistency model.

• •Model availability as date ranges enforced by database exclusion constraints, not application-level check-then-set
• •Separate the search index (eventually consistent, denormalized) from the booking store (strongly consistent, relational)
• •Handle payment with short-lived holds and saga pattern; use idempotency keys to prevent duplicate bookings

Generated with AI, which can make mistakes.