If you have 50,000 servers for your service, and you can reduce that by 1 percent, you save 50 servers. Multiply that by maybe $8k per server and you have saved $400k,you just paid for your self for a year. With meta the numbers are probably a bit bigger.
That's not how it works though. Budgets are annual. A 1% savings of cpu cycles doesn't show up anywhere, it's a rounding error. They don't have a guy that pulls the servers and sells them ahead of the projection. You bought them for 5 years and they're staying. 5 years from now, that 1% got eaten up by other shit.
You're wrong about how services that cost 9+ figures to run annually are budgeted. 1% CPU is absolutely massive and well measured and accounted for in these systems.
You don't buy servers once every 5 years. I've done purchasing every quarter and forecasted a year out. You reduce your services budget for hardware by the amount saved for that year.
According to the video linked somewhere in this thread indicates WhatsApp Erlang workers that want sub-ms latency.
It's definitely for ads auctions
It's Meta. They always push to be that fast on paper, even when it's costly to do and doesn't really need it.
Meta is a humongous company. Any kind of latency has to have a business impact.
If you have 50,000 servers for your service, and you can reduce that by 1 percent, you save 50 servers. Multiply that by maybe $8k per server and you have saved $400k,you just paid for your self for a year. With meta the numbers are probably a bit bigger.
yes, but latency-optimized schedulers tend to have _worse_ throughput, not better.
LOL (I used to work for Meta, so appreciate the facetious understatement)
That's not how it works though. Budgets are annual. A 1% savings of cpu cycles doesn't show up anywhere, it's a rounding error. They don't have a guy that pulls the servers and sells them ahead of the projection. You bought them for 5 years and they're staying. 5 years from now, that 1% got eaten up by other shit.
You're wrong about how services that cost 9+ figures to run annually are budgeted. 1% CPU is absolutely massive and well measured and accounted for in these systems.
2 replies →
You don't buy servers once every 5 years. I've done purchasing every quarter and forecasted a year out. You reduce your services budget for hardware by the amount saved for that year.
2 replies →