Kafka example: “Your food-delivery app just hit 1 million orders a day. Payments, live tracking, notifications - every service is screaming at every other service, and last night the whole thing collapsed for 40 minutes. Somewhere, an order for biryani vanished.”

Notice: no mention of Kafka. Only the pain.

Kafka example: “Now imagine a post office in the middle of the city. Services don’t shout at each other anymore - they drop letters at the post office, and whoever needs them picks them up, at their own pace. Nobody vanishes. That post office is Kafka.”

Kafka example:
1. Producers & consumers - who drops letters, who collects them.
2. Topics - the labeled pigeonholes letters are sorted into.
3. Why it never loses a letter - replication, in one simple picture.

Full Codebasics sample (Kafka, ~85 seconds):

“Your food-delivery app just hit a million orders a day - and last night it collapsed for 40 minutes because every service was shouting at every other service. An order for biryani simply vanished. Now picture a post office in the middle of the city. Services stop shouting; they drop letters at the post office, and whoever needs them collects them at their own pace. That post office is Kafka. Three things make it work. First, producers and consumers - who drops letters, who collects them. Second, topics - labeled pigeonholes the letters are sorted into. Third, replication - the reason no letter is ever lost, even if a whole shelf burns down. And that is why, next time you hit a million orders, nobody’s biryani disappears.”