40 ms·
Maybe its a problem when you’re paying 200ms per small request and you have millions of them ?
by ttoinou 1y ago
Maybe its a problem when you’re paying 200ms per small request and you have millions of them ?
- Eric_WVGG 1y agoSome companies who actually depend on actions like this should consider fronting some cash to keep this project going.
- bastawhiz 1y agoIf you have cold starts on all of your millions of requests, something is very wrong with your code.
- hnlmorg 1y agoThe hypothetical 200ms is for cold starts. That means when the docker container that runs behind the lambda first starts up. When a lambda is invocated, that docker container isn’t immediately destroyed. It’s kept around for reuse. So there isn’t a realistic scenario where you would get a million cold starts in a month.