Struct K8sBackend 

pub struct K8sBackend { /* private fields */ }

Native Kubernetes Lambda execution backend.

Implementations

impl K8sBackend

pub async fn from_env( default_ecr_port: u16, internal_token: String, ) -> Result<Self, K8sBackendError>

Read configuration from env vars and connect to the cluster. Fails fast on missing required config — never silently degrades. default_ecr_port is fakecloud’s bound port; used as the ECR port when FAKECLOUD_K8S_ECR_URL is unset.

Trait Implementations

impl LambdaBackend for K8sBackend

fn reap_stale<'life0, 'async_trait>( &'life0 self, ) -> Pin<Box<dyn Future<Output = ()> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait,

Sweep Lambda Pods that belong to a previous fakecloud process. Without this, a fakecloud restart leaks the previous run’s Pods and Create collides on function names. Mirrors the docker reaper semantics.

fn name(&self) -> &str

Short identifier surfaced via logs and introspection (e.g. "docker", "podman", "kubernetes").

fn launch<'life0, 'life1, 'life2, 'life3, 'life4, 'async_trait>( &'life0 self, func: &'life1 LambdaFunction, _code_zip: Option<&'life2 [u8]>, _layers: &'life3 [Vec<u8>], deploy_id: &'life4 str, ) -> Pin<Box<dyn Future<Output = Result<WarmInstance, RuntimeError>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait, 'life2: 'async_trait, 'life3: 'async_trait, 'life4: 'async_trait,

Launch a fresh runtime instance for func using the supplied code (for zip-package functions) or func.image_uri (for image package functions). layers are the attached layer ZIPs in attach order. deploy_id is the facade-computed fingerprint used to label resources so reaper logic can correlate.

fn terminate<'life0, 'life1, 'async_trait>( &'life0 self, handle: &'life1 BackendHandle, ) -> Pin<Box<dyn Future<Output = ()> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Tear down one instance. Must be idempotent — the facade may call this against an already-gone instance during cleanup races.

fn prepull_image<'life0, 'life1, 'async_trait>( &'life0 self, _image: &'life1 str, ) -> Pin<Box<dyn Future<Output = Result<(), RuntimeError>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Optional hook: pre-warm the runtime image so the first launch() for a function doesn’t pay the cold-pull cost. Called in the background after CreateFunction persists; backends that don’t benefit from pre-pulling (e.g. Kubernetes, which pulls images on the scheduling node anyway) leave this as a no-op.