notion calendar dies when you have at least 4 calendar accounts 😂

While doing research for chapi's 100M rows challenge , i found this neat http load testing tool and its FOSS

https://k6.io/open-source/

https://github.com/grafana/k6

Mongodb could never

this is NOT good for my ssd 😂

I cant even select * from without crashing my pc.

batch insert latency goes up with every additional row and insert rate is sort of plataueing at 100k per sec

fun fact , indexes don't always make your queries faster 😂

also the more cores your pc has ( physical not logical ) , the more workers psql can have on your query at a time

the query returns 2m rows out of the 100m btw

this is so fun , lowkey havent had this much fun learning something in a while

he says at 5 am 😂

i got my query to run at 2.3 s instead of 11s
trying to finish with hardware and config optimizations before i start on shape and index based optimizations

im gonna call it a day now and go gym

day 1 summary :
100M row seed noscript
indexes
system tuning
9 sec optimize on 2m fetch out of 100m rows

tomorrow :
query optimization
shape optimization
partitioning

p95 for a single client 8.3s -> 3.2s ( still horrible but hey a step in the right direction )

someone PLEASE sponsor me 😭
https://github.com/sponsors/RobiMez
jami.bio/robi

lets start tweaking the data structure and think about shape and indexes + what the features will be

this is what im starting with , a heatmap app that tracks how the user uses your app, every x ms the client would basically recocrd the x and y position of the user's mouse , and then record it for that screen.

it would make for a nice vis but i doubt itll actually give value so im just brainstorming and pivoting if needed

Honestly i would love it if i could use this time to rewrite some parts of odit as well , since transactions & sms are technically something that can reach 100m rows.

but its a question of time and scope 😂.