Earlier quoted context omitted.
Today, maybe? I haven't kept up. I'm talking specifically about its launch.
They had 100m account creations within the first 5 days.
Meta’s Hyperscale Infrastructure: Overview and Insights
31–40 of 182 posts
Re: Meta’s Hyperscale Infrastructure: Overview and Insights
#32At least half their gak is due to them NOT moving quickly and NOT wanting to break things.
IIRC, graphql is a means of papering over a bunch of legacy APIs. They removed foreign keys from mysql using it as a column store db, a vestige of the original LAMP stack still on PHP.
I don't think Meta infrastructural choices are applicable to most folk.
What does serverless land your average dev? A high AWS bill. Elastic managed Kubernetes stack? A higher bill.
Did you know that you can use YAML and provision actual cloud provider resources with boring tech? Welcome to Ansible. There is no need to recreate Linux network stack when you have the Linux network stack, and it actually works!
Quite a lot of hacky gak is required when you run node.js as a production public facing web service. A statically compiled binary won't invent novel code execution paths 4 days into a memory leaking runtime bender.
Boring tech is boring, I guess, even if it's new and shiny. Facebook creates tech to mitigate the pathologies their past continuously present.
Re: Meta’s Hyperscale Infrastructure: Overview and Insights
#33I find it interesting how they describe the PHP web front end as a "serverless" or "function as a service" architecture. I guess it's a matter of perspective. It's a service that has a monolithic codebase with lots of endpoints deployed to it. I guess from the perspective of the maintainer of one of those endpoints it's "serverless" but that abstraction (like all abstractions) has leaks: the teams responsible for the…
“Serverless” is not synonymous with Lambda, it’s just a computing model.
Lots of companies are hosting old monoliths on Amazon Fargate, for example.
Re: Meta’s Hyperscale Infrastructure: Overview and Insights
#34Earlier quoted context omitted.
It would be pretty impressive to launch a public cloud I would trust even less than GCP :D
GCP got this reputation because it’s a second class citizen within Google. Google’s own internal infra (Borg, Blaze) is top-notch. If Meta can pull off the public cloud correctly, I’d trust them greatly - they’ve shown significant engineering and product competence till now, even if they could use some more consistent and stable UI.
Dont all new projects go to gcp within google?
Re: Meta’s Hyperscale Infrastructure: Overview and Insights
#35> Moreover, once it [Threads] was developed, the infrastructure teams were given only two day’s notice to prepare for its production launch. Most large organizations would take longer than two days just to draft a project plan involving dozens of interdependent teams, let alone execute it. At Meta, however, we quickly established war rooms across distributed sites, bringing together both infrastructure and product te…
They might have launched it quickly, but nobody (relatively speaking) gave a tinkers damn about the end product. All form and no function.
incumbent networks don't really lose. they saw potential blood in the water at the time with the rumblings of a mass exodus and made an excellent attempt at capitalizing though.
threads as a product was DOA when that didn't work. you need a network of interesting important people for it to be useful. when the migration didn't happen, you ended up with a bunch of instagram meme influencers reposting their content across two apps instead
I think their strategy combined with an open offer exclusivity bonus could have given them the stickiness. up front 5k, 10k, 15k, etc to a twitter user that matches their follower of at least 25k, 50k, 75k, etc count on threads and agrees to exclusively post there for a year. people weren't getting paid on twitter so this would have been alluring
Re: Meta’s Hyperscale Infrastructure: Overview and Insights
#36As much as I hate clownfare, it would be very interesting if they published something like that because of the sheer number of data centers they operate
Re: Meta’s Hyperscale Infrastructure: Overview and Insights
#37> Moreover, once it [Threads] was developed, the infrastructure teams were given only two day’s notice to prepare for its production launch. Most large organizations would take longer than two days just to draft a project plan involving dozens of interdependent teams, let alone execute it. At Meta, however, we quickly established war rooms across distributed sites, bringing together both infrastructure and product te…
They might have launched it quickly, but nobody (relatively speaking) gave a tinkers damn about the end product. All form and no function.
Precisely, for all the talk of efficiency last few years, how do we even begin to measure the total waste of effort and energy of so many smart people that this was? All that effort and stress for effectively nothing, or perhaps even net negative effect.
Re: Meta’s Hyperscale Infrastructure: Overview and Insights
#38Very interesting, in particular the explicit comparisons with hyperscalers. I almost wonder if this is preparation for them launching their own public cloud. Anyone from Meta care to comment?
It would make sense for them to do so if they have variable load but a lot of hardware waiting for action, like Amazon had back when - busy in the evenings, peak load around the holidays, crickets at night. But if they had that, I'm confident they would have been renting out servers a long time ago now. I wonder if they themselves are customers of the cloud providers.
Re: Meta’s Hyperscale Infrastructure: Overview and Insights
#39Earlier quoted context omitted.
GCP got this reputation because it’s a second class citizen within Google. Google’s own internal infra (Borg, Blaze) is top-notch. If Meta can pull off the public cloud correctly, I’d trust them greatly - they’ve shown significant engineering and product competence till now, even if they could use some more consistent and stable UI.
Dont all new projects go to gcp within google?
Yeah at this stage it doesn't make sense to have two tiers within Google.
Re: Meta’s Hyperscale Infrastructure: Overview and Insights
#40Very interesting, in particular the explicit comparisons with hyperscalers. I almost wonder if this is preparation for them launching their own public cloud. Anyone from Meta care to comment?
A lot of the ingredients are there, but it would take a ton of work to separate the Facebook from the internal tools, and make them customer facing.
The internal tools heavily depend on other internal tools, and none of them were written with customers other than Facebook in mind.