Hacker News Reader

Comment by odo1242 | Hacker News Reader

The highest I could think of is WhatsApp, which gets ~1.6 million RPS on average, and they use MQTT

I work in manufacturing, you would be surprised at the volumes of sensor data being processed there for predictive maintenance.

MQTT is everywhere, indeed, because of how simple the protocol is (it’s very ubiquitous in IoT / embedded systems).

DynamoDB scales much more than that. In one of their dynamodb papers they claimed that the amazon us retail website alone made like 89 millions rps during prime day, a few years ago.

It's also one of the most aggravating database systems I've worked with.

It still scales though

89M rps was the peak across all Amazon systems during 2021 Prime Day. Alexa, Amazon.com and the fulfilment centres together, not one site. But its still very impressieve. But the more importarnt number is the per partition limit, 3000 RCU and 1000 WCU. Horizontal scale is the easy part, a table only gets near it if the partition key actually spreads.

WhatsApp is primarily 1 to 1 messaging. Why would this be hard to scale? Scaling horizontally, sharding, and message routing is all it takes, surely?

Wasn't saying it would be hard to scale at all, I just mentioned the highest RPS customer-facing system I could think of off the top of my head.

AWS IAM handles over 1.2 billion requests per second, based on a 2024 post, im sure its higher now https://aws.amazon.com/blogs/security/an-unexpected-discover...

Ah, interesting.

I used to scrape Meta sites at over 200k steady rps for a couple of years, all by myself.

They did send lawyers to my door in the end though.

Someone had forgot IPv6 rate limit bucketing :)

Tell more sounds like a good story

Reply on news.ycombinator.com