Module runtime 

Lambda execution runtime — pluggable backend behind a shared facade.

LambdaRuntime owns the warm-pool bookkeeping and the HTTP invocation path. The actual lifecycle (container/pod create, port discovery, teardown) is delegated to whatever LambdaBackend the facade was constructed with — today that’s DockerBackend; a Kubernetes backend ships in a follow-up.

Re-exports

pub use k8s::K8sBackend;

pub use k8s::K8sBackendError;

Modules

k8s

Kubernetes LambdaBackend implementation.

Structs

DockerBackend

Docker/Podman-based Lambda execution backend.

LambdaRuntime

StreamingInvocation

Wrapper around an in-flight streaming invocation. Yields raw body chunks via Self::next_chunk until the RIE closes the response, at which point the final Ok(None) signals the caller to emit the terminal InvokeComplete frame.

WarmInstance

What LambdaBackend::launch returns. endpoint is the host:port the facade POSTs invocation payloads to; handle is the backend-specific identifier handed back on terminate.

Enums

BackendHandle

Opaque per-backend identifier for a launched runtime instance. The facade hands this back to the backend on teardown so the backend can find the right resource to delete.

RuntimeError

Traits

LambdaBackend

Functions

extract_zip

Extract a ZIP archive to a destination directory.

runtime_to_image

Map AWS runtime identifier to a Docker image tag.

Type Aliases

ContainerRuntime

Backwards-compatible alias used by callers across the workspace (eventbridge bridge, scheduler, lambda_delivery, reset state, server main). Keeping the old name lets the trait extraction land without churn in every consumer.