Hazel WeaklyJun 14, 2024

We ran hachyderm.io, which is ruby on rails + postgres, off of 32 vCPUs, 128GB of ram, and 2 failing hard drives RAID 10'd with international NFS and IOPS performance measuring in the *dozens*

And 45,000 active users

Had the storage been faster than a 1997 floppy disk, we could've cut compute and ram by 80% and not suffered that much at all

Y'all really don't understand how far a single laptop can really push things anymore cause we waste computers so hard

https://mastodon.social/@GeePawHill/112613403362243583

Show threadAndrew Helwer
@hazelweakly one time at $bigtechco I worked on a service that had a custom-written event sourcing architecture deployed on k8s to a cluster of five very expensive cloud VMs. The userbase was 10 people who each used it maybe once every day. It was an internal app. It was slow as shit.
Jun 14, 2024 at 9:10pmWeb
Show thread0xC0DEC0DE07E9Jun 14, 2024
@ahelwer but a database request to another machine (that has the data in RAM already) is about as fast as a fetching the data from a spinning disc! Clearly, we just need to store everything in RAM on remote machines - that’s the lesson, right? @hazelweakly