Documentation ¶
Overview ¶
Package vkubelet implements the core virtual-kubelet framework. It contains everything reuired to implement a virtuak-kubelet, including the core controller which reconciles pod states and API endpoints for things like pod logs, exec, attach, etc.
To get started, call the `New` with the appropriate config. When you are ready to start the controller, which registers the node and starts watching for pod changes, call `Run`. Taints can be used ensure the sceduler only schedules certain workloads to your virtual-kubelet.
vk := vkubelet.New(...) // setup other things ... vk.Run(ctx, ...)
After calling start, cancelling the passed in context will shutdown the controller.
Up to this point you have a running virtual kubelet controller, but no HTTP handlers to deal with requests forwarded from the API server for things like pod logs (e.g. user calls `kubectl logs myVKPod`). This package provides some helpers for this: `AttachPodRoutes` and `AttachMetricsRoutes`.
mux := http.NewServeMux() vkubelet.AttachPodRoutes(provider, mux)
You must configure your own HTTP server, but these helpers will add handlers at the correct URI paths to your serve mux. You are not required to use go's built-in `*http.ServeMux`, but it does implement the `ServeMux` interface defined in this package which is used for these helpers.
Note: The metrics routes may need to be attached to a different HTTP server, depending on your configuration.
For more fine-grained control over the API, see the `vkubelet/api` package which only implements the HTTP handlers that you can use in whatever way you want.
This uses open-cenesus to implement tracing (but no internal metrics yet) which is propagated through the context. This is passed on even to the providers. We may look at supporting custom propagaters for providers who would like to use a different tracing format.
Index ¶
- Constants
- func AttachMetricsRoutes(p providers.Provider, mux ServeMux)
- func AttachPodRoutes(p providers.Provider, mux ServeMux)
- func InstrumentHandler(h http.Handler) http.Handler
- func MetricsSummaryHandler(p providers.Provider) http.Handler
- func NotFound(w http.ResponseWriter, r *http.Request)
- func NotImplemented(w http.ResponseWriter, r *http.Request)
- func PatchNodeStatus(nodes v1.NodeInterface, nodeName types.NodeName, oldNode *corev1.Node, ...) (*corev1.Node, []byte, error)
- func PodHandler(p providers.Provider) http.Handler
- func UpdateNodeLease(ctx context.Context, leases v1beta1.LeaseInterface, lease *coord.Lease) (*coord.Lease, error)
- func UpdateNodeStatus(ctx context.Context, nodes v1.NodeInterface, n *corev1.Node) (*corev1.Node, error)
- type Config
- type NaiveNodeProvider
- type Node
- type NodeOpt
- type PodController
- type ServeMux
- type Server
Constants ¶
const ( // ReasonOptionalConfigMapNotFound is the reason used in events emitted when an optional configmap is not found. ReasonOptionalConfigMapNotFound = "OptionalConfigMapNotFound" // ReasonOptionalConfigMapKeyNotFound is the reason used in events emitted when an optional configmap key is not found. ReasonOptionalConfigMapKeyNotFound = "OptionalConfigMapKeyNotFound" // ReasonFailedToReadOptionalConfigMap is the reason used in events emitted when an optional configmap could not be read. ReasonFailedToReadOptionalConfigMap = "FailedToReadOptionalConfigMap" // ReasonOptionalSecretNotFound is the reason used in events emitted when an optional secret is not found. ReasonOptionalSecretNotFound = "OptionalSecretNotFound" // ReasonOptionalSecretKeyNotFound is the reason used in events emitted when an optional secret key is not found. ReasonOptionalSecretKeyNotFound = "OptionalSecretKeyNotFound" // ReasonFailedToReadOptionalSecret is the reason used in events emitted when an optional secret could not be read. ReasonFailedToReadOptionalSecret = "FailedToReadOptionalSecret" // ReasonMandatoryConfigMapNotFound is the reason used in events emitted when an mandatory configmap is not found. ReasonMandatoryConfigMapNotFound = "MandatoryConfigMapNotFound" // ReasonMandatoryConfigMapKeyNotFound is the reason used in events emitted when an mandatory configmap key is not found. ReasonMandatoryConfigMapKeyNotFound = "MandatoryConfigMapKeyNotFound" // ReasonFailedToReadMandatoryConfigMap is the reason used in events emitted when an mandatory configmap could not be read. ReasonFailedToReadMandatoryConfigMap = "FailedToReadMandatoryConfigMap" // ReasonMandatorySecretNotFound is the reason used in events emitted when an mandatory secret is not found. ReasonMandatorySecretNotFound = "MandatorySecretNotFound" // ReasonMandatorySecretKeyNotFound is the reason used in events emitted when an mandatory secret key is not found. ReasonMandatorySecretKeyNotFound = "MandatorySecretKeyNotFound" // ReasonFailedToReadMandatorySecret is the reason used in events emitted when an mandatory secret could not be read. ReasonFailedToReadMandatorySecret = "FailedToReadMandatorySecret" // ReasonInvalidEnvironmentVariableNames is the reason used in events emitted when a configmap/secret referenced in a ".spec.containers[*].envFrom" field contains invalid environment variable names. ReasonInvalidEnvironmentVariableNames = "InvalidEnvironmentVariableNames" )
const ( DefaultPingInterval = 5 * time.Second DefaultStatusUpdateInterval = 1 * time.Minute )
The default intervals used for lease and status updates.
Variables ¶
This section is empty.
Functions ¶
func AttachMetricsRoutes ¶
AttachMetricsRoutes adds the http routes for pod/node metrics to the passed in serve mux.
Callers should take care to namespace the serve mux as they see fit, however these routes get called by the Kubernetes API server.
func AttachPodRoutes ¶
AttachPodRoutes adds the http routes for pod stuff to the passed in serve mux.
Callers should take care to namespace the serve mux as they see fit, however these routes get called by the Kubernetes API server.
func InstrumentHandler ¶
InstrumentHandler wraps an http.Handler and injects instrumentation into the request context.
func MetricsSummaryHandler ¶
MetricsSummaryHandler creates an http handler for serving pod metrics.
If the passed in provider does not implement providers.PodMetricsProvider, it will create handlers that just serves http.StatusNotImplemented
func NotFound ¶
func NotFound(w http.ResponseWriter, r *http.Request)
NotFound provides a handler for cases where the requested endpoint doesn't exist
func NotImplemented ¶
func NotImplemented(w http.ResponseWriter, r *http.Request)
NotImplemented provides a handler for cases where a provider does not implement a given API
func PatchNodeStatus ¶
func PatchNodeStatus(nodes v1.NodeInterface, nodeName types.NodeName, oldNode *corev1.Node, newNode *corev1.Node) (*corev1.Node, []byte, error)
PatchNodeStatus patches node status. Copied from github.com/kubernetes/kubernetes/pkg/util/node
func PodHandler ¶
PodHandler creates an http handler for interacting with pods/containers.
func UpdateNodeLease ¶
func UpdateNodeLease(ctx context.Context, leases v1beta1.LeaseInterface, lease *coord.Lease) (*coord.Lease, error)
UpdateNodeLease updates the node lease.
If this function returns an errors.IsNotFound(err) error, this likely means that node leases are not supported, if this is the case, call UpdateNodeStatus instead.
If you use this function, it is up to you to syncronize this with other operations.
func UpdateNodeStatus ¶
func UpdateNodeStatus(ctx context.Context, nodes v1.NodeInterface, n *corev1.Node) (*corev1.Node, error)
UpdateNodeStatus triggers an update to the node status in Kubernetes. It first fetches the current node details and then sets the status according to the passed in node object.
If you use this function, it is up to you to syncronize this with other operations. This reduces the time to second-level precision.
Types ¶
type Config ¶
type Config struct { Client *kubernetes.Clientset Namespace string NodeName string Provider providers.Provider ResourceManager *manager.ResourceManager PodSyncWorkers int PodInformer corev1informers.PodInformer }
Config is used to configure a new server.
type NaiveNodeProvider ¶
type NaiveNodeProvider struct{}
NaiveNodeProvider is a basic node provider that only uses the passed in context on `Ping` to determine if the node is healthy.
func (NaiveNodeProvider) NotifyNodeStatus ¶
func (NaiveNodeProvider) NotifyNodeStatus(ctx context.Context, f func(*corev1.Node))
NotifyNodeStatus implements the NodeProvider interface.
This NaiveNodeProvider does not support updating node status and so this function is a no-op.
type Node ¶
type Node struct {
// contains filtered or unexported fields
}
Node deals with creating and managing a node object in Kubernetes. It can register a node with Kubernetes and periodically update its status.
func NewNode ¶
func NewNode(p providers.NodeProvider, node *corev1.Node, leases v1beta1.LeaseInterface, nodes v1.NodeInterface, opts ...NodeOpt) (*Node, error)
NewNode creates a new node. This does not have any side-effects on the system or kubernetes.
Use the node's `Run` method to register and run the loops to update the node in Kubernetes.
func (*Node) Run ¶
Run registers the node in kubernetes and starts loops for updating the node status in Kubernetes.
The node status must be updated periodically in Kubertnetes to keep the node active. Newer versions of Kubernetes support node leases, which are essentially light weight pings. Older versions of Kubernetes require updating the node status periodically.
If Kubernetes supports node leases this will use leases with a much slower node status update (because some things still expect the node to be updated periodically), otherwise it will only use node status update with the configured ping interval.
type NodeOpt ¶
NodeOpt are the functional options used for configuring a node
func WithNodeDisableLease ¶
WithNodeDisableLease forces node leases to be disabled and to only update using node status Note that this will force status updates to occur on the ping interval frequency
func WithNodeLease ¶
WithNodeLease sets the base node lease to use. If a lease time is set, it will be ignored.
func WithNodePingInterval ¶
WithNodePingInterval sets the inteval for checking node status
func WithNodeStatusUpdateInterval ¶
WithNodeStatusUpdateInterval sets the interval for updating node status This is only used when leases are supported and only for updating the actual node status, not the node lease.
type PodController ¶
type PodController struct {
// contains filtered or unexported fields
}
PodController is the controller implementation for Pod resources.
func NewPodController ¶
func NewPodController(server *Server) *PodController
NewPodController returns a new instance of PodController.
func (*PodController) Run ¶
func (pc *PodController) Run(ctx context.Context, threadiness int) error
Run will set up the event handlers for types we are interested in, as well as syncing informer caches and starting workers. It will block until stopCh is closed, at which point it will shutdown the work queue and wait for workers to finish processing their current work items.
type ServeMux ¶
ServeMux defines an interface used to attach routes to an existing http serve mux. It is used to enable callers creating a new server to completely manage their own HTTP server while allowing us to attach the required routes to satisfy the Kubelet HTTP interfaces.
type Server ¶
type Server struct {
// contains filtered or unexported fields
}
Server masquarades itself as a kubelet and allows for the virtual node to be backed by non-vm/node providers.
func New ¶
New creates a new virtual-kubelet server. This is the entrypoint to this package.
This creates but does not start the server. You must call `Run` on the returned object to start the server.
func (*Server) Ready ¶
func (s *Server) Ready() <-chan struct{}
Ready returns a channel which will be closed once the VKubelet is running
func (*Server) Run ¶
Run creates and starts an instance of the pod controller, blocking until it stops.
Note that this does not setup the HTTP routes that are used to expose pod info to the Kubernetes API Server, such as logs, metrics, exec, etc. See `AttachPodRoutes` and `AttachMetricsRoutes` to set these up.