Starting a new engineering job? Before you propose replacing the system that looks obviously wrong, find out why the team built it that way.
Queue ride requests to absorb bursts, then scale matching workers. The queue buys resilience, but waiting time still matters: accepting requests faster than you can match them creates unhappy riders.
1
146
Two riders can request the same driver. Use a shared lock with an expiry while making an offer, and move on if the driver doesn’t respond. A lock inside one server won’t coordinate the others.
1
144
Let the phone adapt location updates to movement. A parked driver doesn’t need to send GPS coordinates as often as one moving through a city. Fewer writes, with a freshness tradeoff.
1
1
238
Driver locations change constantly. Keep the latest positions in a geospatial store such as Redis so finding nearby drivers doesn’t require scanning a database table.
1
1
341
System design question:
A seller updates a price. The product page shows it immediately, but search still shows the old price, sometimes forever.
Products live in Postgres. Search uses a separate index.
What could cause this, and how would you fix it?
Cache event details and use Elasticsearch for discovery. Treat the seat map as a hint: availability can change between viewing a seat and trying to reserve it.
1
2
783
A virtual waiting room controls how many people reach checkout. More booking servers alone won’t fix millions of users competing for the same few seats.
1
3
1,362
Reserve seats for a few minutes while the user pays. An expiring Redis lock avoids holding a database transaction open through checkout. The final booking still needs a database concurrency check.
1
4
1,714











