kube-rs

Rust client for Kubernetes in the style of a more generic client-go plus a runtime abstraction inspired by controller-runtime.

These crate makes certain assumptions about the kubernetes api to allow writing generic abstractions, and as such contains rust reinterpretations of reflectors, informers, and controller so that you can writing kubernetes controllers/watchers/operators more easily.

NB: This library is currently undergoing a lot of changes with async/await stabilizing. Please check the CHANGELOG when upgrading.

Installation

Select a version of kube along with the generated k8s api types that corresponds to your cluster version:

[dependencies]
kube = "0.37.0"
kube-derive = "0.37.0"
kube-runtime = "0.37.0"
k8s-openapi = { version = "0.9.0", default-features = false, features = ["v1_17"] }

Note that turning off default-features for k8s-openapi is recommended to speed up your compilation (and we provide an api anyway).

Usage

See the examples directory for how to watch over resources in a simplistic way.

API Docs

Some real world examples:

• version-rs: super lightweight reflector deployment with actix 2 and prometheus metrics

• controller-rs: Controller owned by a Manager inside actix

• krustlet: a complete WASM running kubelet

Api

The direct Api type takes a client, and is constructed with either the ::global or ::namespaced functions:

use k8s_openapi::api::core::v1::Pod;
let pods: Api<Pod> = Api::namespaced(client, "default");

let p = pods.get("blog").await?;
println!("Got blog pod with containers: {:?}", p.spec.unwrap().containers);

let patch = json!({"spec": {
    "activeDeadlineSeconds": 5
}});
let patched = pods.patch("blog", &pp, serde_json::to_vec(&patch)?).await?;
assert_eq!(patched.spec.active_deadline_seconds, Some(5));

pods.delete("blog", &DeleteParams::default()).await?;

See the examples ending in _api examples for more detail.

Custom Resource Definitions

Working with custom resources uses automatic code-generation via proc_macros in kube-derive.

You need to #[derive(CustomResource)] and some #[kube(attrs..)] on a spec struct:

#[derive(CustomResource, Serialize, Deserialize, Default, Clone)]
#[kube(group = "clux.dev", version = "v1", namespaced)]
pub struct FooSpec {
    name: String,
    info: String,
}

Then you can use a lot of generated code as:

println!("kind = {}", Foo::KIND); // impl k8s_openapi::Resource
let foos: Api<Foo> = Api::namespaced(client, "default");
let f = Foo::new("my-foo");
println!("foo: {:?}", f)
println!("crd: {}", serde_yaml::to_string(Foo::crd());

There are a ton of kubebuilder like instructions that you can annotate with here. See the crd_ prefixed examples for more.

Runtime

The kube_runtime create contains sets of higher level abstractions on top of the Api and Resource types so that you don't have to do all the watch/resourceVersion/storage book-keeping yourself.

Watchers

A low level streaming interface (similar to informers) that presents Applied, Deleted or Restarted events.

let api = Api::<Pod>::namespaced(client, "default");
let watcher = watcher(api, ListParams::default());

This now gives a continual stream of events and you do not need to care about the watch having to restart, or connections dropping.

let apply_events = try_flatten_applied(watcher).boxed_local()
while let Some(event) = watcher.try_next().await? {
    println!("Applied: {}", Meta::name(&event));
}

NB: the plain stream items a watcher returns are different from WatchEvent. If you are following along to "see what changed", you should flatten it with one of the utilities like try_flatten_applied or try_flatten_touched.

Reflectors

A reflector is a watcher with Store on K. It acts on all the Event<K> exposed by watcher to ensure that the state in the Store is as accurate as possible.

let nodes: Api<Node> = Api::namespaced(client, &namespace);
let lp = ListParams::default()
    .labels("beta.kubernetes.io/instance-type=m4.2xlarge");
let store = reflector::store::Writer::<Node>::default();
let reader = store.as_reader();
let rf = reflector(store, watcher(nodes, lp));

At this point you can listen to the reflector as if it was a watcher, but you can also query the reader at any point.

Controllers

A Controller is a reflector along with an arbitrary number of watchers that schedule events internally to send events through a reconciler:

Controller::new(root_kind_api, ListParams::default())
    .owns(child_kind_api, ListParams::default())
    .run(reconcile, error_policy, context)
    .for_each(|res| async move {
        match res {
            Ok(o) => info!("reconciled {:?}", o),
            Err(e) => warn!("reconcile failed: {}", Report::from(e)),
        }
    })
    .await;

Here reconcile and error_policy refer to functions you define. The first will be called when the root or child elements change, and the second when the reconciler returns an Err.

Examples

Examples that show a little common flows. These all have logging of this library set up to debug, and where possible pick up on the NAMESPACE evar.

NB: not all examples have been migrated to the new runtime yet. If it uses kube-runtime it's new.

# watch configmap events
cargo run --example configmap_watcher
# watch pod events
cargo run --example pod_informer
# watch event events
cargo run --example event_informer
# watch for broken nodes
cargo run --example node_informer

or for the reflectors:

cargo run --example pod_reflector
cargo run --example node_reflector
cargo run --example deployment_reflector
cargo run --example secret_reflector
cargo run --example configmap_reflector

for one based on a CRD, you need to create the CRD first:

kubectl apply -f examples/foo.yaml
cargo run --example crd_reflector

then you can kubectl apply -f crd-baz.yaml -n default, or kubectl delete -f crd-baz.yaml -n default, or kubectl edit foos baz -n default to verify that the events are being picked up.

ditto for a controller:

kubectl apply -f kube/examples/configmapgen_controller_crd.yaml
cargo run --example configmapgen_controller &
kubectl apply -f kube/examples/configmapgen_controller_object.yaml

For straight API use examples, try:

cargo run --example crd_api
cargo run --example job_api
cargo run --example log_stream
cargo run --example pod_api
NAMESPACE=dev cargo run --example log_stream -- kafka-manager-7d4f4bd8dc-f6c44

Rustls

Kube has basic support (with caveats) for rustls as a replacement for the openssl dependency. To use this, turn off default features, and enable rustls-tls:

cargo run --example pod_informer --no-default-features --features=rustls-tls

or in Cargo.toml:

[dependencies]
kube = { version = "0.37.0", default-features = false, features = ["rustls-tls"] }
kube-runtime = { version = "0.37.0", default-features = false, features = ["rustls-tls"] }
k8s-openapi = { version = "0.9.0", default-features = false, features = ["v1_17"] }

This will pull in the variant of reqwest that also uses its rustls-tls feature.

License

Apache 2.0 licensed. See LICENSE for details.