Yep, vertical scaling goes a long way. But it’s not compute where the bottleneck for scale lies, rather in the resiliency & availability. So although a single

Yep, vertical scaling goes a long way. But it’s not compute where the bottleneck for scale lies, rather in the resiliency & availability. So although a single server goes a long way, to hit that sweet 99.999 SLA, people horizontally scales way before hitting the maximum compute capacity of a singe machine. HA makes everything way more difficult to operate and reason about.