3 ms·
Remember you don’t pay for invocations. You pay for memory, latency and traffic. That constrains how you build architectures with lambda. Anything bloated or w
by uuyi 5y ago
Remember you don’t pay for invocations. You pay for memory, latency and traffic.
That constrains how you build architectures with lambda. Anything bloated or written in low efficiency languages, or does any amount of compute or latent network traffic will burn you. That includes dealing with slow clients and synchronous APIs.
As for rules, there aren’t really any catch-all ones. We build out on Kubernetes now though as that’s portable. The lambda developer story is quite horrible and that’s where our cost is.
- 015a 5y agoI'm fairly certain you do pay per invocation; $0.20/million. Its not the bulk of the cost for sure but its non-zero nonetheless.
- uuyi 5y agoYes that’s the point. The marketing points to that as the big cost but it’s insignificant compared to the real cost which is running workloads on it that aren’t optimised for price first. Which is probably most of them from experience. Then again that’s how Bezos rolls.