We’ve been working to reduce the base costs of running our cloud infrastructure on AWS by switching to more efficient instance types for our core nodes. This is the core infrastructure we use to ensure hubs are “always available” for users, even when no one is actively using a hub. By moving from older r5.xlarge instances to newer, more efficient r8i-flex.large instances, we’ve significantly reduced these baseline costs while maintaining the same level of service. Here’s a plot of d...
