Kubernetes API Concepts

The Kubernetes API is a resource-based (RESTful) programmatic interface provided via HTTP. It supports retrieving, creating, updating, and deleting primary resources via the standard HTTP verbs (POST, PUT, PATCH, DELETE, GET).

For some resources, the API includes additional subresources that allow fine grained authorization (such as separate views for Pod details and log retrievals), and can accept and serve those resources in different representations for convenience or efficiency.

Kubernetes supports efficient change notifications on resources via watches. Kubernetes also provides consistent list operations so that API clients can effectively cache, track, and synchronize the state of resources.

You can view the API reference online, or read on to learn about the API in general.

Kubernetes API terminology

Kubernetes generally leverages common RESTful terminology to describe the API concepts:

• A resource type is the name used in the URL (pods, namespaces, services)
• All resource types have a concrete representation (their object schema) which is called a kind
• A list of instances of a resource type is known as a collection
• A single instance of a resource type is called a resource, and also usually represents an object
• For some resource types, the API includes one or more sub-resources, which are represented as URI paths below the resource

Most Kubernetes API resource types are objects – they represent a concrete instance of a concept on the cluster, like a pod or namespace. A smaller number of API resource types are virtual in that they often represent operations on objects, rather than objects, such as a permission check (use a POST with a JSON-encoded body of SubjectAccessReview to the subjectaccessreviews resource), or the eviction sub-resource of a Pod (used to trigger API-initiated eviction).

Object names

All objects you can create via the API have a unique object name to allow idempotent creation and retrieval, except that virtual resource types may not have unique names if they are not retrievable, or do not rely on idempotency. Within a namespace, only one object of a given kind can have a given name at a time. However, if you delete the object, you can make a new object with the same name. Some objects are not namespaced (for example: Nodes), and so their names must be unique across the whole cluster.

API verbs

Almost all object resource types support the standard HTTP verbs - GET, POST, PUT, PATCH, and DELETE. Kubernetes also uses its own verbs, which are often written lowercase to distinguish them from HTTP verbs.

Kubernetes uses the term list to describe returning a collection of resources to distinguish from retrieving a single resource which is usually called a get. If you sent an HTTP GET request with the ?watch query parameter, Kubernetes calls this a watch and not a get (see Efficient detection of changes for more details).

For PUT requests, Kubernetes internally classifies these as either create or update based on the state of the existing object. An update is different from a patch; the HTTP verb for a patch is PATCH.

Resource URIs

All resource types are either scoped by the cluster (/apis/GROUP/VERSION/*) or to a namespace (/apis/GROUP/VERSION/namespaces/NAMESPACE/*). A namespace-scoped resource type will be deleted when its namespace is deleted and access to that resource type is controlled by authorization checks on the namespace scope.

Note: core resources use /api instead of /apis and omit the GROUP path segment.

Examples:

• /api/v1/namespaces
• /api/v1/pods
• /api/v1/namespaces/my-namespace/pods
• /apis/apps/v1/deployments
• /apis/apps/v1/namespaces/my-namespace/deployments
• /apis/apps/v1/namespaces/my-namespace/deployments/my-deployment

You can also access collections of resources (for example: listing all Nodes). The following paths are used to retrieve collections and resources:

• Cluster-scoped resources:

  • GET /apis/GROUP/VERSION/RESOURCETYPE - return the collection of resources of the resource type
  • GET /apis/GROUP/VERSION/RESOURCETYPE/NAME - return the resource with NAME under the resource type

• Namespace-scoped resources:

  • GET /apis/GROUP/VERSION/RESOURCETYPE - return the collection of all instances of the resource type across all namespaces
  • GET /apis/GROUP/VERSION/namespaces/NAMESPACE/RESOURCETYPE - return collection of all instances of the resource type in NAMESPACE
  • GET /apis/GROUP/VERSION/namespaces/NAMESPACE/RESOURCETYPE/NAME - return the instance of the resource type with NAME in NAMESPACE

Since a namespace is a cluster-scoped resource type, you can retrieve the list (“collection”) of all namespaces with GET /api/v1/namespaces and details about a particular namespace with GET /api/v1/namespaces/NAME.

• Cluster-scoped subresource: GET /apis/GROUP/VERSION/RESOURCETYPE/NAME/SUBRESOURCE
• Namespace-scoped subresource: GET /apis/GROUP/VERSION/namespaces/NAMESPACE/RESOURCETYPE/NAME/SUBRESOURCE

The verbs supported for each subresource will differ depending on the object - see the API reference for more information. It is not possible to access sub-resources across multiple resources - generally a new virtual resource type would be used if that becomes necessary.

Efficient detection of changes

The Kubernetes API allows clients to make an initial request for an object or a collection, and then to track changes since that initial request: a watch. Clients can send a list or a get and then make a follow-up watch request.

To make this change tracking possible, every Kubernetes object has a resourceVersion field representing the version of that resource as stored in the underlying persistence layer. When retrieving a collection of resources (either namespace or cluster scoped), the response from the API server contains a resourceVersion value. The client can use that resourceVersion to initiate a watch against the API server.

When you send a watch request, the API server responds with a stream of changes. These changes itemize the outcome of operations (such as create, delete, and update) that occurred after the resourceVersion you specified as a parameter to the watch request. The overall watch mechanism allows a client to fetch the current state and then subscribe to subsequent changes, without missing any events.

If a client watch is disconnected then that client can start a new watch from the last returned resourceVersion; the client could also perform a fresh get / list request and begin again. See Resource Version Semantics for more detail.

For example:

1. List all of the pods in a given namespace.

   GET /api/v1/namespaces/test/pods
   ---
   200 OK
   Content-Type: application/json
   
   {
     "kind": "PodList",
     "apiVersion": "v1",
     "metadata": {"resourceVersion":"10245"},
     "items": [...]
   }
   

2. Starting from resource version 10245, receive notifications of any API operations (such as create, delete, patch or update) that affect Pods in the test namespace. Each change notification is a JSON document. The HTTP response body (served as application/json) consists a series of JSON documents.

   GET /api/v1/namespaces/test/pods?watch=1&resourceVersion=10245
   ---
   200 OK
   Transfer-Encoding: chunked
   Content-Type: application/json
   
   {
     "type": "ADDED",
     "object": {"kind": "Pod", "apiVersion": "v1", "metadata": {"resourceVersion": "10596", ...}, ...}
   }
   {
     "type": "MODIFIED",
     "object": {"kind": "Pod", "apiVersion": "v1", "metadata": {"resourceVersion": "11020", ...}, ...}
   }
   ...
   

A given Kubernetes server will only preserve a historical record of changes for a limited time. Clusters using etcd 3 preserve changes in the last 5 minutes by default. When the requested watch operations fail because the historical version of that resource is not available, clients must handle the case by recognizing the status code 410 Gone, clearing their local cache, performing a new get or list operation, and starting the watch from the resourceVersion that was returned.

For subscribing to collections, Kubernetes client libraries typically offer some form of standard tool for this list-then-watch logic. (In the Go client library, this is called a Reflector and is located in the k8s.io/client-go/tools/cache package.)

Watch bookmarks

To mitigate the impact of short history window, the Kubernetes API provides a watch event named BOOKMARK. It is a special kind of event to mark that all changes up to a given resourceVersion the client is requesting have already been sent. The document representing the BOOKMARK event is of the type requested by the request, but only includes a .metadata.resourceVersion field. For example:

GET /api/v1/namespaces/test/pods?watch=1&resourceVersion=10245&allowWatchBookmarks=true
---
200 OK
Transfer-Encoding: chunked
Content-Type: application/json

{
  "type": "ADDED",
  "object": {"kind": "Pod", "apiVersion": "v1", "metadata": {"resourceVersion": "10596", ...}, ...}
}
...
{
  "type": "BOOKMARK",
  "object": {"kind": "Pod", "apiVersion": "v1", "metadata": {"resourceVersion": "12746"} }
}

As a client, you can request BOOKMARK events by setting the allowWatchBookmarks=true query parameter to a watch request, but you shouldn't assume bookmarks are returned at any specific interval, nor can clients assume that the API server will send any BOOKMARK event even when requested.

Streaming lists

On large clusters, retrieving the collection of some resource types may result in a significant increase of resource usage (primarily RAM) on the control plane. In order to alleviate its impact and simplify the user experience of the list + watch pattern, Kubernetes v1.27 introduces as an alpha feature the support for requesting the initial state (previously requested via the list request) as part of the watch request.

Provided that the WatchList feature gate is enabled, this can be achieved by specifying sendInitialEvents=true as query string parameter in a watch request. If set, the API server starts the watch stream with synthetic init events (of type ADDED) to build the whole state of all existing objects followed by a BOOKMARK event (if requested via allowWatchBookmarks=true option). The bookmark event includes the resource version to which is synced. After sending the bookmark event, the API server continues as for any other watch request.

When you set sendInitialEvents=true in the query string, Kubernetes also requires that you set resourceVersionMatch to NotOlderThan value. If you provided resourceVersion in the query string without providing a value or don't provide it at all, this is interpreted as a request for consistent read; the bookmark event is sent when the state is synced at least to the moment of a consistent read from when the request started to be processed. If you specify resourceVersion (in the query string), the bookmark event is sent when the state is synced at least to the provided resource version.

Example

An example: you want to watch a collection of Pods. For that collection, the current resource version is 10245 and there are two pods: foo and bar. Then sending the following request (explicitly requesting consistent read by setting empty resource version using resourceVersion=) could result in the following sequence of events:

GET /api/v1/namespaces/test/pods?watch=1&sendInitialEvents=true&allowWatchBookmarks=true&resourceVersion=&resourceVersionMatch=NotOlderThan
---
200 OK
Transfer-Encoding: chunked
Content-Type: application/json

{
  "type": "ADDED",
  "object": {"kind": "Pod", "apiVersion": "v1", "metadata": {"resourceVersion": "8467", "name": "foo"}, ...}
}
{
  "type": "ADDED",
  "object": {"kind": "Pod", "apiVersion": "v1", "metadata": {"resourceVersion": "5726", "name": "bar"}, ...}
}
{
  "type": "BOOKMARK",
  "object": {"kind": "Pod", "apiVersion": "v1", "metadata": {"resourceVersion": "10245"} }
}
...
<followed by regular watch stream starting from resourceVersion="10245">

Response compression

APIResponseCompression is an option that allows the API server to compress the responses for get and list requests, reducing the network bandwidth and improving the performance of large-scale clusters. It is enabled by default since Kubernetes 1.16 and it can be disabled by including APIResponseCompression=false in the --feature-gates flag on the API server.

API response compression can significantly reduce the size of the response, especially for large resources or collections. For example, a list request for pods can return hundreds of kilobytes or even megabytes of data, depending on the number of pods and their attributes. By compressing the response, the network bandwidth can be saved and the latency can be reduced.

To verify if APIResponseCompression is working, you can send a get or list request to the API server with an Accept-Encoding header, and check the response size and headers. For example:

GET /api/v1/pods
Accept-Encoding: gzip
---
200 OK
Content-Type: application/json
content-encoding: gzip
...

The content-encoding header indicates that the response is compressed with gzip.

Retrieving large results sets in chunks

On large clusters, retrieving the collection of some resource types may result in very large responses that can impact the server and client. For instance, a cluster may have tens of thousands of Pods, each of which is equivalent to roughly 2 KiB of encoded JSON. Retrieving all pods across all namespaces may result in a very large response (10-20MB) and consume a large amount of server resources.

The Kubernetes API server supports the ability to break a single large collection request into many smaller chunks while preserving the consistency of the total request. Each chunk can be returned sequentially which reduces both the total size of the request and allows user-oriented clients to display results incrementally to improve responsiveness.

You can request that the API server handles a list by serving single collection using pages (which Kubernetes calls chunks). To retrieve a single collection in chunks, two query parameters limit and continue are supported on requests against collections, and a response field continue is returned from all list operations in the collection's metadata field. A client should specify the maximum results they wish to receive in each chunk with limit and the server will return up to limit resources in the result and include a continue value if there are more resources in the collection.

As an API client, you can then pass this continue value to the API server on the next request, to instruct the server to return the next page (chunk) of results. By continuing until the server returns an empty continue value, you can retrieve the entire collection.

Like a watch operation, a continue token will expire after a short amount of time (by default 5 minutes) and return a 410 Gone if more results cannot be returned. In this case, the client will need to start from the beginning or omit the limit parameter.

For example, if there are 1,253 pods on the cluster and you want to receive chunks of 500 pods at a time, request those chunks as follows:

1. List all of the pods on a cluster, retrieving up to 500 pods each time.

   GET /api/v1/pods?limit=500
   ---
   200 OK
   Content-Type: application/json
   
   {
     "kind": "PodList",
     "apiVersion": "v1",
     "metadata": {
       "resourceVersion":"10245",
       "continue": "ENCODED_CONTINUE_TOKEN",
       "remainingItemCount": 753,
       ...
     },
     "items": [...] // returns pods 1-500
   }
   

2. Continue the previous call, retrieving the next set of 500 pods.

   GET /api/v1/pods?limit=500&continue=ENCODED_CONTINUE_TOKEN
   ---
   200 OK
   Content-Type: application/json
   
   {
     "kind": "PodList",
     "apiVersion": "v1",
     "metadata": {
       "resourceVersion":"10245",
       "continue": "ENCODED_CONTINUE_TOKEN_2",
       "remainingItemCount": 253,
       ...
     },
     "items": [...] // returns pods 501-1000
   }
   

3. Continue the previous call, retrieving the last 253 pods.

   GET /api/v1/pods?limit=500&continue=ENCODED_CONTINUE_TOKEN_2
   ---
   200 OK
   Content-Type: application/json
   
   {
     "kind": "PodList",
     "apiVersion": "v1",
     "metadata": {
       "resourceVersion":"10245",
       "continue": "", // continue token is empty because we have reached the end of the list
       ...
     },
     "items": [...] // returns pods 1001-1253
   }
   

Notice that the resourceVersion of the collection remains constant across each request, indicating the server is showing you a consistent snapshot of the pods. Pods that are created, updated, or deleted after version 10245 would not be shown unless you make a separate list request without the continue token. This allows you to break large requests into smaller chunks and then perform a watch operation on the full set without missing any updates.

remainingItemCount is the number of subsequent items in the collection that are not included in this response. If the list request contained label or field selectors then the number of remaining items is unknown and the API server does not include a remainingItemCount field in its response. If the list is complete (either because it is not chunking, or because this is the last chunk), then there are no more remaining items and the API server does not include a remainingItemCount field in its response. The intended use of the remainingItemCount is estimating the size of a collection.

Collections

In Kubernetes terminology, the response you get from a list is a collection. However, Kubernetes defines concrete kinds for collections of different types of resource. Collections have a kind named for the resource kind, with List appended.

When you query the API for a particular type, all items returned by that query are of that type. For example, when you list Services, the collection response has kind set to ServiceList; each item in that collection represents a single Service. For example:

GET /api/v1/services

{
  "kind": "ServiceList",
  "apiVersion": "v1",
  "metadata": {
    "resourceVersion": "2947301"
  },
  "items": [
    {
      "metadata": {
        "name": "kubernetes",
        "namespace": "default",
...
      "metadata": {
        "name": "kube-dns",
        "namespace": "kube-system",
...

There are dozens of collection types (such as PodList, ServiceList, and NodeList) defined in the Kubernetes API. You can get more information about each collection type from the Kubernetes API documentation.

Some tools, such as kubectl, represent the Kubernetes collection mechanism slightly differently from the Kubernetes API itself. Because the output of kubectl might include the response from multiple list operations at the API level, kubectl represents a list of items using kind: List. For example:

kubectl get services -A -o yaml

apiVersion: v1
kind: List
metadata:
  resourceVersion: ""
  selfLink: ""
items:
- apiVersion: v1
  kind: Service
  metadata:
    creationTimestamp: "2021-06-03T14:54:12Z"
    labels:
      component: apiserver
      provider: kubernetes
    name: kubernetes
    namespace: default
...
- apiVersion: v1
  kind: Service
  metadata:
    annotations:
      prometheus.io/port: "9153"
      prometheus.io/scrape: "true"
    creationTimestamp: "2021-06-03T14:54:14Z"
    labels:
      k8s-app: kube-dns
      kubernetes.io/cluster-service: "true"
      kubernetes.io/name: CoreDNS
    name: kube-dns
    namespace: kube-system

Receiving resources as Tables

When you run kubectl get, the default output format is a simple tabular representation of one or more instances of a particular resource type. In the past, clients were required to reproduce the tabular and describe output implemented in kubectl to perform simple lists of objects. A few limitations of that approach include non-trivial logic when dealing with certain objects. Additionally, types provided by API aggregation or third party resources are not known at compile time. This means that generic implementations had to be in place for types unrecognized by a client.

In order to avoid potential limitations as described above, clients may request the Table representation of objects, delegating specific details of printing to the server. The Kubernetes API implements standard HTTP content type negotiation: passing an Accept header containing a value of application/json;as=Table;g=meta.k8s.io;v=v1 with a GET call will request that the server return objects in the Table content type.

For example, list all of the pods on a cluster in the Table format.

GET /api/v1/pods
Accept: application/json;as=Table;g=meta.k8s.io;v=v1
---
200 OK
Content-Type: application/json

{
    "kind": "Table",
    "apiVersion": "meta.k8s.io/v1",
    ...
    "columnDefinitions": [
        ...
    ]
}

For API resource types that do not have a custom Table definition known to the control plane, the API server returns a default Table response that consists of the resource's name and creationTimestamp fields.

GET /apis/crd.example.com/v1alpha1/namespaces/default/resources
---
200 OK
Content-Type: application/json
...

{
    "kind": "Table",
    "apiVersion": "meta.k8s.io/v1",
    ...
    "columnDefinitions": [
        {
            "name": "Name",
            "type": "string",
            ...
        },
        {
            "name": "Created At",
            "type": "date",
            ...
        }
    ]
}

Not all API resource types support a Table response; for example, a CustomResourceDefinitions might not define field-to-table mappings, and an APIService that extends the core Kubernetes API might not serve Table responses at all. If you are implementing a client that uses the Table information and must work against all resource types, including extensions, you should make requests that specify multiple content types in the Accept header. For example:

Accept: application/json;as=Table;g=meta.k8s.io;v=v1, application/json

Alternate representations of resources

By default, Kubernetes returns objects serialized to JSON with content type application/json. This is the default serialization format for the API. However, clients may request the more efficient Protobuf representation of these objects for better performance at scale. The Kubernetes API implements standard HTTP content type negotiation: passing an Accept header with a GET call will request that the server tries to return a response in your preferred media type, while sending an object in Protobuf to the server for a PUT or POST call means that you must set the Content-Type header appropriately.

The server will return a response with a Content-Type header if the requested format is supported, or the 406 Not acceptable error if none of the media types you requested are supported. All built-in resource types support the application/json media type.

See the Kubernetes API reference for a list of supported content types for each API.

For example:

1. List all of the pods on a cluster in Protobuf format.

   GET /api/v1/pods
   Accept: application/vnd.kubernetes.protobuf
   ---
   200 OK
   Content-Type: application/vnd.kubernetes.protobuf
   
   ... binary encoded PodList object
   

2. Create a pod by sending Protobuf encoded data to the server, but request a response in JSON.

   POST /api/v1/namespaces/test/pods
   Content-Type: application/vnd.kubernetes.protobuf
   Accept: application/json
   ... binary encoded Pod object
   ---
   200 OK
   Content-Type: application/json
   
   {
     "kind": "Pod",
     "apiVersion": "v1",
     ...
   }
   

Not all API resource types support Protobuf; specifically, Protobuf isn't available for resources that are defined as CustomResourceDefinitions or are served via the aggregation layer. As a client, if you might need to work with extension types you should specify multiple content types in the request Accept header to support fallback to JSON. For example:

Accept: application/vnd.kubernetes.protobuf, application/json

Kubernetes Protobuf encoding

Kubernetes uses an envelope wrapper to encode Protobuf responses. That wrapper starts with a 4 byte magic number to help identify content in disk or in etcd as Protobuf (as opposed to JSON), and then is followed by a Protobuf encoded wrapper message, which describes the encoding and type of the underlying object and then contains the object.

The wrapper format is:

A four byte magic number prefix:
  Bytes 0-3: "k8s\x00" [0x6b, 0x38, 0x73, 0x00]

An encoded Protobuf message with the following IDL:
  message Unknown {
    // typeMeta should have the string values for "kind" and "apiVersion" as set on the JSON object
    optional TypeMeta typeMeta = 1;

    // raw will hold the complete serialized object in protobuf. See the protobuf definitions in the client libraries for a given kind.
    optional bytes raw = 2;

    // contentEncoding is encoding used for the raw data. Unspecified means no encoding.
    optional string contentEncoding = 3;

    // contentType is the serialization method used to serialize 'raw'. Unspecified means application/vnd.kubernetes.protobuf and is usually
    // omitted.
    optional string contentType = 4;
  }

  message TypeMeta {
    // apiVersion is the group/version for this type
    optional string apiVersion = 1;
    // kind is the name of the object schema. A protobuf definition should exist for this object.
    optional string kind = 2;
  }

Resource deletion

When you delete a resource this takes place in two phases.

1. finalization
2. removal

{
  "kind": "ConfigMap",
  "apiVersion": "v1",
  "metadata": {
    "finalizers": ["url.io/neat-finalization", "other-url.io/my-finalizer"],
    "deletionTimestamp": nil,
  }
}

When a client first sends a delete to request the removal of a resource, the .metadata.deletionTimestamp is set to the current time. Once the .metadata.deletionTimestamp is set, external controllers that act on finalizers may start performing their cleanup work at any time, in any order.

Order is not enforced between finalizers because it would introduce significant risk of stuck .metadata.finalizers.

The .metadata.finalizers field is shared: any actor with permission can reorder it. If the finalizer list were processed in order, then this might lead to a situation in which the component responsible for the first finalizer in the list is waiting for some signal (field value, external system, or other) produced by a component responsible for a finalizer later in the list, resulting in a deadlock.

Without enforced ordering, finalizers are free to order amongst themselves and are not vulnerable to ordering changes in the list.

Once the last finalizer is removed, the resource is actually removed from etcd.

Single resource API

The Kubernetes API verbs get, create, update, patch, delete and proxy support single resources only. These verbs with single resource support have no support for submitting multiple resources together in an ordered or unordered list or transaction.

When clients (including kubectl) act on a set of resources, the client makes a series of single-resource API requests, then aggregates the responses if needed.

By contrast, the Kubernetes API verbs list and watch allow getting multiple resources, and deletecollection allows deleting multiple resources.

Field validation

Kubernetes always validates the type of fields. For example, if a field in the API is defined as a number, you cannot set the field to a text value. If a field is defined as an array of strings, you can only provide an array. Some fields allow you to omit them, other fields are required. Omitting a required field from an API request is an error.

If you make a request with an extra field, one that the cluster's control plane does not recognize, then the behavior of the API server is more complicated.

By default, the API server drops fields that it does not recognize from an input that it receives (for example, the JSON body of a PUT request).

There are two situations where the API server drops fields that you supplied in an HTTP request.

These situations are:

1. The field is unrecognized because it is not in the resource's OpenAPI schema. (One exception to this is for CRDs that explicitly choose not to prune unknown fields via x-kubernetes-preserve-unknown-fields).
2. The field is duplicated in the object.

Validation for unrecognized or duplicate fields

From 1.25 onward, unrecognized or duplicate fields in an object are detected via validation on the server when you use HTTP verbs that can submit data (POST, PUT, and PATCH). Possible levels of validation are Ignore, Warn (default), and Strict.

Ignore

The API server succeeds in handling the request as it would without the erroneous fields being set, dropping all unknown and duplicate fields and giving no indication it has done so.

Warn

(Default) The API server succeeds in handling the request, and reports a warning to the client. The warning is sent using the Warning: response header, adding one warning item for each unknown or duplicate field. For more information about warnings and the Kubernetes API, see the blog article Warning: Helpful Warnings Ahead.

Strict

The API server rejects the request with a 400 Bad Request error when it detects any unknown or duplicate fields. The response message from the API server specifies all the unknown or duplicate fields that the API server has detected.

The field validation level is set by the fieldValidation query parameter.

Tools that submit requests to the server (such as kubectl), might set their own defaults that are different from the Warn validation level that the API server uses by default.

The kubectl tool uses the --validate flag to set the level of field validation. It accepts the values ignore, warn, and strict while also accepting the values true (equivalent to strict) and false (equivalent to ignore). The default validation setting for kubectl is --validate=true, which means strict server-side field validation.

When kubectl cannot connect to an API server with field validation (API servers prior to Kubernetes 1.27), it will fall back to using client-side validation. Client-side validation will be removed entirely in a future version of kubectl.

Dry-run

When you use HTTP verbs that can modify resources (POST, PUT, PATCH, and DELETE), you can submit your request in a dry run mode. Dry run mode helps to evaluate a request through the typical request stages (admission chain, validation, merge conflicts) up until persisting objects to storage. The response body for the request is as close as possible to a non-dry-run response. Kubernetes guarantees that dry-run requests will not be persisted in storage or have any other side effects.

Make a dry-run request

Dry-run is triggered by setting the dryRun query parameter. This parameter is a string, working as an enum, and the only accepted values are:

[no value set]

Allow side effects. You request this with a query string such as ?dryRun or ?dryRun&pretty=true. The response is the final object that would have been persisted, or an error if the request could not be fulfilled.

All

Every stage runs as normal, except for the final storage stage where side effects are prevented.

When you set ?dryRun=All, any relevant admission controllers are run, validating admission controllers check the request post-mutation, merge is performed on PATCH, fields are defaulted, and schema validation occurs. The changes are not persisted to the underlying storage, but the final object which would have been persisted is still returned to the user, along with the normal status code.

If the non-dry-run version of a request would trigger an admission controller that has side effects, the request will be failed rather than risk an unwanted side effect. All built in admission control plugins support dry-run. Additionally, admission webhooks can declare in their configuration object that they do not have side effects, by setting their sideEffects field to None.

Here is an example dry-run request that uses ?dryRun=All:

POST /api/v1/namespaces/test/pods?dryRun=All
Content-Type: application/json
Accept: application/json

The response would look the same as for non-dry-run request, but the values of some generated fields may differ.

Generated values

Some values of an object are typically generated before the object is persisted. It is important not to rely upon the values of these fields set by a dry-run request, since these values will likely be different in dry-run mode from when the real request is made. Some of these fields are:

• name: if generateName is set, name will have a unique random name
• creationTimestamp / deletionTimestamp: records the time of creation/deletion
• UID: uniquely identifies the object and is randomly generated (non-deterministic)
• resourceVersion: tracks the persisted version of the object
• Any field set by a mutating admission controller
• For the Service resource: Ports or IP addresses that the kube-apiserver assigns to Service objects

Dry-run authorization

Authorization for dry-run and non-dry-run requests is identical. Thus, to make a dry-run request, you must be authorized to make the non-dry-run request.

For example, to run a dry-run patch for a Deployment, you must be authorized to perform that patch. Here is an example of a rule for Kubernetes RBAC that allows patching Deployments:

rules:
- apiGroups: ["apps"]
  resources: ["deployments"]
  verbs: ["patch"]

See Authorization Overview.

Updates to existing resources

Kubernetes provides several ways to update existing objects. You can read choosing an update mechanism to learn about which approach might be best for your use case.

You can overwrite (update) an existing resource - for example, a ConfigMap - using an HTTP PUT. For a PUT request, it is the client's responsibility to specify the resourceVersion (taking this from the object being updated). Kubernetes uses that resourceVersion information so that the API server can detect lost updates and reject requests made by a client that is out of date with the cluster. In the event that the resource has changed (the resourceVersion the client provided is stale), the API server returns a 409 Conflict error response.

Instead of sending a PUT request, the client can send an instruction to the API server to patch an existing resource. A patch is typically appropriate if the change that the client wants to make isn't conditional on the existing data. Clients that need effective detection of lost updates should consider making their request conditional on the existing resourceVersion (either HTTP PUT or HTTP PATCH), and then handle any retries that are needed in case there is a conflict.

The Kubernetes API supports four different PATCH operations, determined by their corresponding HTTP Content-Type header:

application/apply-patch+yaml

Server Side Apply YAML (a Kubernetes-specific extension, based on YAML). All JSON documents are valid YAML, so you can also submit JSON using this media type. See Server Side Apply serialization for more details.
To Kubernetes, this is a create operation if the object does not exist, or a patch operation if the object already exists.

application/json-patch+json

JSON Patch, as defined in RFC6902. A JSON patch is a sequence of operations that are executed on the resource; for example {"op": "add", "path": "/a/b/c", "value": [ "foo", "bar" ]}.
To Kubernetes, this is a patch operation.

A patch using application/json-patch+json can include conditions to validate consistency, allowing the operation to fail if those conditions are not met (for example, to avoid a lost update).

application/merge-patch+json

JSON Merge Patch, as defined in RFC7386. A JSON Merge Patch is essentially a partial representation of the resource. The submitted JSON is combined with the current resource to create a new one, then the new one is saved.
To Kubernetes, this is a patch operation.

application/strategic-merge-patch+json

Strategic Merge Patch (a Kubernetes-specific extension based on JSON). Strategic Merge Patch is a custom implementation of JSON Merge Patch. You can only use Strategic Merge Patch with built-in APIs, or with aggregated API servers that have special support for it. You cannot use application/strategic-merge-patch+json with any API defined using a CustomResourceDefinition.

Note:

The Kubernetes server side apply mechanism has superseded Strategic Merge Patch.

Kubernetes' Server Side Apply feature allows the control plane to track managed fields for newly created objects. Server Side Apply provides a clear pattern for managing field conflicts, offers server-side apply and update operations, and replaces the client-side functionality of kubectl apply.

For Server-Side Apply, Kubernetes treats the request as a create if the object does not yet exist, and a patch otherwise. For other requests that use PATCH at the HTTP level, the logical Kubernetes operation is always patch.

See Server Side Apply for more details.

Choosing an update mechanism

HTTP PUT to replace existing resource

The update (HTTP PUT) operation is simple to implement and flexible, but has drawbacks:

• You need to handle conflicts where the resourceVersion of the object changes between your client reading it and trying to write it back. Kubernetes always detects the conflict, but you as the client author need to implement retries.
• You might accidentally drop fields if you decode an object locally (for example, using client-go, you could receive fields that your client does not know how to handle - and then drop them as part of your update.
• If there's a lot of contention on the object (even on a field, or set of fields, that you're not trying to edit), you might have trouble sending the update. The problem is worse for larger objects and for objects with many fields.

HTTP PATCH using JSON Patch

A patch update is helpful, because:

• As you're only sending differences, you have less data to send in the PATCH request.
• You can make changes that rely on existing values, such as copying the value of a particular field into an annotation.
• Unlike with an update (HTTP PUT), making your change can happen right away even if there are frequent changes to unrelated fields): you usually would not need to retry.
  • You might still need to specify the resourceVersion (to match an existing object) if you want to be extra careful to avoid lost updates
  • It's still good practice to write in some retry logic in case of errors.
• You can use test conditions to careful craft specific update conditions. For example, you can increment a counter without reading it if the existing value matches what you expect. You can do this with no lost update risk, even if the object has changed in other ways since you last wrote to it. (If the test condition fails, you can fall back to reading the current value and then write back the changed number).

However:

• you need more local (client) logic to build the patch; it helps a lot if you have a library implementation of JSON Patch, or even for making a JSON Patch specifically against Kubernetes
• as the author of client software, you need to be careful when building the patch (the HTTP request body) not to drop fields (the order of operations matters)

HTTP PATCH using Server-Side Apply

Server-Side Apply has some clear benefits:

• A single round trip: it rarely requires making a GET request first.
  • and you can still detect conflicts for unexpected changes
  • you have the option to force override a conflict, if appropriate
• Client implementations are easy to make
• You get an atomic create-or-update operation without extra effort (similar to UPSERT in some SQL dialects)

However:

• Server-Side Apply does not work at all for field changes that depend on a current value of the object
• You can only apply updates to objects. Some resources in the Kubernetes HTTP API are not objects (they do not have a .metadata field), and Server-Side Apply is only relevant for Kubernetes objects.

Resource versions

Resource versions are strings that identify the server's internal version of an object. Resource versions can be used by clients to determine when objects have changed, or to express data consistency requirements when getting, listing and watching resources. Resource versions must be treated as opaque by clients and passed unmodified back to the server.

You must not assume resource versions are numeric or collatable. API clients may only compare two resource versions for equality (this means that you must not compare resource versions for greater-than or less-than relationships).

resourceVersion fields in metadata

Clients find resource versions in resources, including the resources from the response stream for a watch, or when using list to enumerate resources.

v1.meta/ObjectMeta - The metadata.resourceVersion of a resource instance identifies the resource version the instance was last modified at.

v1.meta/ListMeta - The metadata.resourceVersion of a resource collection (the response to a list) identifies the resource version at which the collection was constructed.

resourceVersion parameters in query strings

The get, list, and watch operations support the resourceVersion parameter. From version v1.19, Kubernetes API servers also support the resourceVersionMatch parameter on list requests.

The API server interprets the resourceVersion parameter differently depending on the operation you request, and on the value of resourceVersion. If you set resourceVersionMatch then this also affects the way matching happens.

Semantics for get and list

For get and list, the semantics of resourceVersion are:

get:

list:

From version v1.19, Kubernetes API servers support the resourceVersionMatch parameter on list requests. If you set both resourceVersion and resourceVersionMatch, the resourceVersionMatch parameter determines how the API server interprets resourceVersion.

You should always set the resourceVersionMatch parameter when setting resourceVersion on a list request. However, be prepared to handle the case where the API server that responds is unaware of resourceVersionMatch and ignores it.

Unless you have strong consistency requirements, using resourceVersionMatch=NotOlderThan and a known resourceVersion is preferable since it can achieve better performance and scalability of your cluster than leaving resourceVersion and resourceVersionMatch unset, which requires quorum read to be served.

Setting the resourceVersionMatch parameter without setting resourceVersion is not valid.

This table explains the behavior of list requests with various combinations of resourceVersion and resourceVersionMatch:

The meaning of the get and list semantics are:

Any

Return data at any resource version. The newest available resource version is preferred, but strong consistency is not required; data at any resource version may be served. It is possible for the request to return data at a much older resource version that the client has previously observed, particularly in high availability configurations, due to partitions or stale caches. Clients that cannot tolerate this should not use this semantic.

Most recent

Return data at the most recent resource version. The returned data must be consistent (in detail: served from etcd via a quorum read).

Not older than

Return data at least as new as the provided resourceVersion. The newest available data is preferred, but any data not older than the provided resourceVersion may be served. For list requests to servers that honor the resourceVersionMatch parameter, this guarantees that the collection's .metadata.resourceVersion is not older than the requested resourceVersion, but does not make any guarantee about the .metadata.resourceVersion of any of the items in that collection.

Exact

Return data at the exact resource version provided. If the provided resourceVersion is unavailable, the server responds with HTTP 410 "Gone". For list requests to servers that honor the resourceVersionMatch parameter, this guarantees that the collection's .metadata.resourceVersion is the same as the resourceVersion you requested in the query string. That guarantee does not apply to the .metadata.resourceVersion of any items within that collection.

Continue Token, Exact

Return data at the resource version of the initial paginated list call. The returned continue tokens are responsible for keeping track of the initially provided resource version for all paginated list calls after the initial paginated list.

When using resourceVersionMatch=NotOlderThan and limit is set, clients must handle HTTP 410 "Gone" responses. For example, the client might retry with a newer resourceVersion or fall back to resourceVersion="".

When using resourceVersionMatch=Exact and limit is unset, clients must verify that the collection's .metadata.resourceVersion matches the requested resourceVersion, and handle the case where it does not. For example, the client might fall back to a request with limit set.

Semantics for watch

For watch, the semantics of resource version are:

watch:

The meaning of those watch semantics are:

Get State and Start at Any

Caution:

Watches initialized this way may return arbitrarily stale data. Please review this semantic before using it, and favor the other semantics where possible.

Start a watch at any resource version; the most recent resource version available is preferred, but not required. Any starting resource version is allowed. It is possible for the watch to start at a much older resource version that the client has previously observed, particularly in high availability configurations, due to partitions or stale caches. Clients that cannot tolerate this apparent rewinding should not start a watch with this semantic. To establish initial state, the watch begins with synthetic "Added" events for all resource instances that exist at the starting resource version. All following watch events are for all changes that occurred after the resource version the watch started at.

Get State and Start at Most Recent

Start a watch at the most recent resource version, which must be consistent (in detail: served from etcd via a quorum read). To establish initial state, the watch begins with synthetic "Added" events of all resources instances that exist at the starting resource version. All following watch events are for all changes that occurred after the resource version the watch started at.

Start at Exact

Start a watch at an exact resource version. The watch events are for all changes after the provided resource version. Unlike "Get State and Start at Most Recent" and "Get State and Start at Any", the watch is not started with synthetic "Added" events for the provided resource version. The client is assumed to already have the initial state at the starting resource version since the client provided the resource version.

"410 Gone" responses

Servers are not required to serve all older resource versions and may return a HTTP 410 (Gone) status code if a client requests a resourceVersion older than the server has retained. Clients must be able to tolerate 410 (Gone) responses. See Efficient detection of changes for details on how to handle 410 (Gone) responses when watching resources.

If you request a resourceVersion outside the applicable limit then, depending on whether a request is served from cache or not, the API server may reply with a 410 Gone HTTP response.

Unavailable resource versions

Servers are not required to serve unrecognized resource versions. If you request list or get for a resource version that the API server does not recognize, then the API server may either:

• wait briefly for the resource version to become available, then timeout with a 504 (Gateway Timeout) if the provided resource versions does not become available in a reasonable amount of time;
• respond with a Retry-After response header indicating how many seconds a client should wait before retrying the request.

If you request a resource version that an API server does not recognize, the kube-apiserver additionally identifies its error responses with a "Too large resource version" message.

If you make a watch request for an unrecognized resource version, the API server may wait indefinitely (until the request timeout) for the resource version to become available.