All terms
Glossary
AWS Lambda
The service that runs individual functions without your own servers, triggered by events or calls.
Billing is per invocation and per runtime times allocated memory. More memory also means more CPU, which is why a function with double the memory sometimes costs less, because it finishes sooner.
Typical uses are event processing, scheduled jobs and interfaces with strongly varying load.
How you notice it
- There are short tasks reacting to events.
- Running a dedicated instance would not pay off.
- Load arrives in waves.
Frequently asked
How do I find the right memory setting?
By measuring, not guessing. Run the same function at 256, 512 and 1024 MB and compare duration times price. Often the larger setting is both faster and cheaper, because CPU scales with memory.