From 77b4856e800bef7d65948560dff66ab17bd4ff41 Mon Sep 17 00:00:00 2001 From: Tim Bannister Date: Thu, 8 Aug 2019 08:53:51 +0100 Subject: [PATCH] Add controller concept --- content/en/docs/concepts/_index.md | 10 +- .../concepts/architecture/cloud-controller.md | 2 +- .../docs/concepts/architecture/controller.md | 160 ++++++++++++++++++ .../en/docs/reference/glossary/controller.md | 7 +- 4 files changed, 167 insertions(+), 12 deletions(-) create mode 100644 content/en/docs/concepts/architecture/controller.md diff --git a/content/en/docs/concepts/_index.md b/content/en/docs/concepts/_index.md index 8608c9600c64c..991b5552ce34a 100644 --- a/content/en/docs/concepts/_index.md +++ b/content/en/docs/concepts/_index.md @@ -17,7 +17,7 @@ The Concepts section helps you learn about the parts of the Kubernetes system an To work with Kubernetes, you use *Kubernetes API objects* to describe your cluster's *desired state*: what applications or other workloads you want to run, what container images they use, the number of replicas, what network and disk resources you want to make available, and more. You set your desired state by creating objects using the Kubernetes API, typically via the command-line interface, `kubectl`. You can also use the Kubernetes API directly to interact with the cluster and set or modify your desired state. -Once you've set your desired state, the *Kubernetes Control Plane* makes the cluster's current state match the desired state via the Pod Lifecycle Event Generator (PLEG). To do so, Kubernetes performs a variety of tasks automatically--such as starting or restarting containers, scaling the number of replicas of a given application, and more. The Kubernetes Control Plane consists of a collection of processes running on your cluster: +Once you've set your desired state, the *Kubernetes Control Plane* makes the cluster's current state match the desired state via the Pod Lifecycle Event Generator (PLEG). To do so, Kubernetes performs a variety of tasks automatically--such as starting or restarting containers, scaling the number of replicas of a given application, and more. The Kubernetes Control Plane consists of a collection of processes running on your cluster: * The **Kubernetes Master** is a collection of three processes that run on a single node in your cluster, which is designated as the master node. Those processes are: [kube-apiserver](/docs/admin/kube-apiserver/), [kube-controller-manager](/docs/admin/kube-controller-manager/) and [kube-scheduler](/docs/admin/kube-scheduler/). * Each individual non-master node in your cluster runs two processes: @@ -26,7 +26,7 @@ Once you've set your desired state, the *Kubernetes Control Plane* makes the clu ## Kubernetes Objects -Kubernetes contains a number of abstractions that represent the state of your system: deployed containerized applications and workloads, their associated network and disk resources, and other information about what your cluster is doing. These abstractions are represented by objects in the Kubernetes API; see the [Kubernetes Objects overview](/docs/concepts/abstractions/overview/) for more details. +Kubernetes contains a number of abstractions that represent the state of your system: deployed containerized applications and workloads, their associated network and disk resources, and other information about what your cluster is doing. These abstractions are represented by objects in the Kubernetes API. See [Understanding Kubernetes Objects](/docs/concepts/overview/working-with-objects/kubernetes-objects/) for more details. The basic Kubernetes objects include: @@ -35,12 +35,12 @@ The basic Kubernetes objects include: * [Volume](/docs/concepts/storage/volumes/) * [Namespace](/docs/concepts/overview/working-with-objects/namespaces/) -In addition, Kubernetes contains a number of higher-level abstractions called Controllers. Controllers build upon the basic objects, and provide additional functionality and convenience features. They include: +Kubernetes also contains higher-level abstractions that rely on [Controllers](/docs/concepts/architecture/controller/) to build upon the basic objects, and provide additional functionality and convenience features. These include: -* [ReplicaSet](/docs/concepts/workloads/controllers/replicaset/) * [Deployment](/docs/concepts/workloads/controllers/deployment/) -* [StatefulSet](/docs/concepts/workloads/controllers/statefulset/) * [DaemonSet](/docs/concepts/workloads/controllers/daemonset/) +* [StatefulSet](/docs/concepts/workloads/controllers/statefulset/) +* [ReplicaSet](/docs/concepts/workloads/controllers/replicaset/) * [Job](/docs/concepts/workloads/controllers/jobs-run-to-completion/) ## Kubernetes Control Plane diff --git a/content/en/docs/concepts/architecture/cloud-controller.md b/content/en/docs/concepts/architecture/cloud-controller.md index 1c62bf741c26a..acee852bc2a42 100644 --- a/content/en/docs/concepts/architecture/cloud-controller.md +++ b/content/en/docs/concepts/architecture/cloud-controller.md @@ -1,7 +1,7 @@ --- title: Concepts Underlying the Cloud Controller Manager content_template: templates/concept -weight: 30 +weight: 40 --- {{% capture overview %}} diff --git a/content/en/docs/concepts/architecture/controller.md b/content/en/docs/concepts/architecture/controller.md new file mode 100644 index 0000000000000..9a80f1414dd1f --- /dev/null +++ b/content/en/docs/concepts/architecture/controller.md @@ -0,0 +1,160 @@ +--- +title: Controllers +content_template: templates/concept +weight: 30 +--- + +{{% capture overview %}} + +In robotics and automation, a _control loop_ is +a non-terminating loop that regulates the state of a system. + +Here is one example of a control loop: a thermostat in a room. + +When you set the temperature, that's telling the thermostat +about your *desired state*. The actual room temperature is the +*current state*. The thermostat acts to bring the current state +closer to the desired state, by turning equipment on or off. + +In Kubernetes, a _controller_ is a control loop that watches the +state of your +{{< glossary_tooltip term_id="cluster" text="cluster">}}, then +makes or requests changes. Each controller tries to move the current +cluster state closer to the desired state. + +{{% /capture %}} + + +{{% capture body %}} + +## Controller pattern + +A controller tracks at least one Kubernetes resource type. +These [objects](/docs/concepts/overview/working-with-objects/kubernetes-objects/) +and their associated data are the desired state, so the +controller(s) for that resource are responsible for making the current +state come closer to that desired state. + +The controller might carry the action out itself; more commonly, in Kubernetes, +a controller will send messages to the +{{< glossary_tooltip text="API server" term_id="kube-apiserver" >}} that have +useful side effects. + +### Control via API server + +The controller for Job is an example of a built-in controller that +makes all of its changes by interacting with your cluster's API server. + +{{< glossary_tooltip term_id="job" >}} is a Kubernetes resource that +runs a {{< glossary_tooltip term_id="pod" >}}, or perhaps several Pods, +to carry out a task and then stop. + +(Once [scheduled](/docs/concepts/scheduling/), Pod objects become part +of the desired state for a kubelet). + +When the Job controller sees a new task it makes sure that, somewhere +in your cluster, the kubelets on a set of Nodes are running the right +number of Pods to get the work done. +The Job controller does not run any Pods or containers +itself. Instead, the Job controller tells the API server to create or remove +Pod [objects](/docs/concepts/overview/working-with-objects/kubernetes-objects/). +Other components in the +{{< glossary_tooltip text="control plane" term_id="control-plane" >}} +act on the new information (there are new Pods to schedule and run), +and eventually the work is done. + +After you create a new Job, the desired state is for that Job to +be completed. The Job controller makes the current state be nearer +to the desired state, by creating Pods that do the work you +wanted. + +Controllers also update the objects that configure them +For example: once the work is done for a Job, the Job controller +updates that Job object to mark it `Finished`. + +(This is a bit like how some thermostats turn a light off to +indicate that the your room is now at the temperature you set). + +### Direct control + +By contrast with Job, some controllers need to make changes to +things outside of your cluster. + +For example, if you use a control loop to make sure there +are enough {{< glossary_tooltip text="Nodes" term_id="node" >}} +in your cluster, then that controller needs something outside the +current cluster to set up new Nodes when needed. + +This kind of controller finds its desired state from the API +server and then communicates directly with an external system +to bring the current state closer in line. + +(There actually is a controller that horizontally scales the +nodes in your cluster. See +[Cluster autoscaling](https://kubernetes.io/docs/tasks/administer-cluster/cluster-management/#cluster-autoscaling)). + +## Desired versus current state {#desired-vs-current} + +Kubernetes takes a cloud-native view of systems, and is able to handle +constant change. + +Your cluster could be changing at any point as work happens and +control loops automatically fix failures. This means that, +potentially, your cluster never reaches a stable state. + +So long as the controllers for your cluster are making useful changes, +it's OK for the current state to be different from the desired state +at a particular moment. + +## Design + +As a tenet of its design, Kubernetes uses lots of controllers that each manage +a particular aspect of cluster state. Most commonly, a particular control loop +(controller) uses one kind of resource as its desired state, and has a different +kind of resource that it manages to make that desired state happen. + +It's useful to have simple controllers rather than one, monolithic set of control +loops that are interlinked. Controllers can fail, so Kubernetes is designed to +allow for that. + +For example: a controller for Jobs tracks Job objects (to discover +new work) and Pod object (to run the Jobs, and then to see when the work is +finished). In this case something else creates the Job objects, whereas the Job +controller creates some Pod objects. + +{{< note >}} +There can be several controllers that create or update the same kind of object. +Behind the scenes, Kubernetes controllers make sure that they only pay attention +to the resources linked to their controlling resource. + +For example, you can have Deployments and Jobs; these both create Pods. +The Job controller does not delete the Pods that your Deployment created, +because there is information ({{< glossary_tooltip term_id="label" text="labels" >}}) +the controllers can use to tell those Pods apart. +{{< /note >}} + +## Ways of running controllers {#running-controllers} + +Kubernetes comes with a set of built-in controllers that run inside +the {{< glossary_tooltip term_id="kube-controller-manager" >}}. These +built-in controllers provide important core behaviors. + +The Deployment controller and Job controller are examples of controllers that +come as part of Kubernetes itself (“built-in” controllers). +Kubernetes lets you run a resilient control plane, so that if any of the built-in +controllers were to fail, another part of the control plane will take over the work. + +You can find controllers that run outside the control plane, to extend Kubernetes. +Or, if you want, you can write a new controller yourself. +You can run your own controller as a set of Pods, +or externally to Kubernetes. What fits best will depend on what that particular +controller does. + +{{% /capture %}} + +{{% capture whatsnext %}} +* Read about the [Kubernetes control plane](https://kubernetes.io/docs/concepts/#kubernetes-control-plane) +* Discover some of the basic [Kubernetes objects](https://kubernetes.io/docs/concepts/#kubernetes-objects) +* Learn more about the [Kubernetes API](/docs/concepts/overview/kubernetes-api/) +* If you want to write your own controller, see [Extension Patterns](/docs/concepts/extend-kubernetes/extend-cluster/#extension-patterns) in Extending Kubernetes. +{{% /capture %}} diff --git a/content/en/docs/reference/glossary/controller.md b/content/en/docs/reference/glossary/controller.md index 5c810e2b47280..af34a24a5a549 100755 --- a/content/en/docs/reference/glossary/controller.md +++ b/content/en/docs/reference/glossary/controller.md @@ -2,7 +2,7 @@ title: Controller id: controller date: 2018-04-12 -full_link: /docs/admin/kube-controller-manager/ +full_link: /docs/concepts/architecture/controller/ short_description: > A control loop that watches the shared state of the cluster through the apiserver and makes changes attempting to move the current state towards the desired state. @@ -12,8 +12,3 @@ tags: - fundamental --- A control loop that watches the shared state of the cluster through the {{< glossary_tooltip text="apiserver" term_id="kube-apiserver" >}} and makes changes attempting to move the current state towards the desired state. - - - -Examples of controllers that ship with Kubernetes today are the replication controller, endpoints controller, namespace controller, and serviceaccounts controller. -