What you declare for a function
The stack entry is the function name. Source defaults to functions/{name}. Environment variables and cron schedules can sit on that entry. The gateway invokes the function when a route targets it. Compute is 0.01 credits per GiB-second, and an idle function does not keep a replica charge.
Reviewers see the function in the plan next to any route that calls it. Deploy publishes the built bundle. Rollback returns to the previous plan. The handler is not wrapped in a separate API product and a separate log group.
What stays behind in Lambda
A Lambda configuration is not the function entry. Reserved concurrency, layers, file systems, destinations, and event-source mappings are not read from an AWS account and written into stack.json.
You bring Go source and a name. You do not export the Lambda and expect its runtime settings to reappear. Node, Python, and PHP belong on workers, which are container processes, when the job does not fit a Go WebAssembly handler.
AWS Lambda vs KERNSTACK on European Bare Metal
| Dimension | AWS Lambda | KERNSTACK Functions |
|---|---|---|
| Operator | AWS account | KERNSTACK project |
| Manifest | AWS Lambda console or API | resources.functions |
| Region | AWS region list | European bare metal |
| Bill | AWS Lambda line item | One project credit balance |
Frequently asked questions
No. You declare a name under resources.functions and ship Go source. Memory, layers, and event-source mappings are not copied.
Go compiled to WebAssembly. Container jobs in Node.js, Python, PHP, or a Go binary are workers, not functions.
Nothing for compute. Compute is 0.01 credits per GiB-second only while the handler runs.
Related capabilities, architectures, and comparisons
Ship your stack on the Software Lifecycle Cloud.
Closed preview on dedicated European infrastructure. Bring a repository or start from a blueprint.