←back to thread

Postgres LISTEN/NOTIFY does not scale

(www.recall.ai)
570 points davidgu | 1 comments | 07 Jul 25 14:05 UTC | HN request time: 0.235s | source
Show context
callamdelaney ◴[10 Jul 25 23:18 UTC] No.44526788[source]▶
My kneejerk reaction to the headline is ‘why would it?’.

It’s unsurprising to me that an AI company appears to have chosen exactly the wrong tool for the job.

replies(3): >>44526797 #>>44527255 #>>44528010 #
kristianc ◴[11 Jul 25 00:34 UTC] No.44527255[source]▶
Sounds like a deliberate attempt to avoid spinning up Redis, Kafka, or an outbox system early on.. and then underestimated how quickly their scale would make it blow up. Story as old as time.
replies(4): >>44527509 #>>44527547 #>>44529439 #>>44529665 #
const_cast ◴[11 Jul 25 01:26 UTC] No.44527547[source]▶
I find the opposite story more true: additional complexity in the form of caching early, for a scale that never comes. I've worked on one too many sprawling, distributed systems with too little users to justify it.
replies(1): >>44528682 #
physix ◴[11 Jul 25 05:32 UTC] No.44528682[source]▶
"Sprawling distributed systems".

I like that. Sounds like a synonym for "Platform Engineering". :-)

I remember being amazed that lambda architecture was considered a kind of reference, when it looked to me more like a workaround.

We like to build IT cathedrals, until we have to run them.

replies(1): >>44529620 #
1. const_cast ◴[11 Jul 25 08:25 UTC] No.44529620[source]▶
If there's one thing I took away from school, it's that distributed systems are hard. More failure points and much more communication hops. Serialization into deserialization into serialization again over network hops.