Merge pull request #130447 from pohly/dra-device-taints

device taints and tolerations (KEP 5055)

api/discovery/aggregated_v2.json

@@ -1915,6 +1915,26 @@
                 "watch"
               ]
             },
+            {
+              "resource": "devicetaintrules",
+              "responseKind": {
+                "group": "",
+                "kind": "DeviceTaintRule",
+                "version": ""
+              },
+              "scope": "Cluster",
+              "singularResource": "devicetaintrule",
+              "verbs": [
+                "create",
+                "delete",
+                "deletecollection",
+                "get",
+                "list",
+                "patch",
+                "update",
+                "watch"
+              ]
+            },
             {
               "resource": "resourceclaims",
               "responseKind": {

api/discovery/apis__resource.k8s.io__v1alpha3.json

@@ -20,6 +20,23 @@
         "watch"
       ]
     },
+    {
+      "kind": "DeviceTaintRule",
+      "name": "devicetaintrules",
+      "namespaced": false,
+      "singularName": "devicetaintrule",
+      "storageVersionHash": "DJ3UJ0fj8MI=",
+      "verbs": [
+        "create",
+        "delete",
+        "deletecollection",
+        "get",
+        "list",
+        "patch",
+        "update",
+        "watch"
+      ]
+    },
     {
       "kind": "ResourceClaim",
       "name": "resourceclaims",

api/openapi-spec/v3/apis__resource.k8s.io__v1beta1_openapi.json

@@ -193,6 +193,19 @@
             },
             "description": "Capacity defines the set of capacities for this device. The name of each capacity must be unique in that set.\n\nThe maximum number of attributes and capacities combined is 32.",
             "type": "object"
+          },
+          "taints": {
+            "description": "If specified, these are the driver-defined taints.\n\nThe maximum number of taints is 8.\n\nThis is an alpha field and requires enabling the DRADeviceTaints feature gate.",
+            "items": {
+              "allOf": [
+                {
+                  "$ref": "#/components/schemas/io.k8s.api.resource.v1beta1.DeviceTaint"
+                }
+              ],
+              "default": {}
+            },
+            "type": "array",
+            "x-kubernetes-list-type": "atomic"
           }
         },
         "type": "object"
@@ -607,6 +620,19 @@
             },
             "type": "array",
             "x-kubernetes-list-type": "atomic"
+          },
+          "tolerations": {
+            "description": "If specified, the request's tolerations.\n\nTolerations for NoSchedule are required to allocate a device which has a taint with that effect. The same applies to NoExecute.\n\nIn addition, should any of the allocated devices get tainted with NoExecute after allocation and that effect is not tolerated, then all pods consuming the ResourceClaim get deleted to evict them. The scheduler will not let new pods reserve the claim while it has these tainted devices. Once all pods are evicted, the claim will get deallocated.\n\nThe maximum number of tolerations is 16.\n\nThis field can only be set when deviceClassName is set and no subrequests are specified in the firstAvailable list.\n\nThis is an alpha field and requires enabling the DRADeviceTaints feature gate.",
+            "items": {
+              "allOf": [
+                {
+                  "$ref": "#/components/schemas/io.k8s.api.resource.v1beta1.DeviceToleration"
+                }
+              ],
+              "default": {}
+            },
+            "type": "array",
+            "x-kubernetes-list-type": "atomic"
           }
         },
         "required": [
@@ -640,6 +666,19 @@
             "default": "",
             "description": "Request is the name of the request in the claim which caused this device to be allocated. If it references a subrequest in the firstAvailable list on a DeviceRequest, this field must include both the name of the main request and the subrequest using the format <main request>/<subrequest>.\n\nMultiple devices may have been allocated per request.",
             "type": "string"
+          },
+          "tolerations": {
+            "description": "A copy of all tolerations specified in the request at the time when the device got allocated.\n\nThe maximum number of tolerations is 16.\n\nThis is an alpha field and requires enabling the DRADeviceTaints feature gate.",
+            "items": {
+              "allOf": [
+                {
+                  "$ref": "#/components/schemas/io.k8s.api.resource.v1beta1.DeviceToleration"
+                }
+              ],
+              "default": {}
+            },
+            "type": "array",
+            "x-kubernetes-list-type": "atomic"
           }
         },
         "required": [
@@ -698,6 +737,19 @@
             },
             "type": "array",
             "x-kubernetes-list-type": "atomic"
+          },
+          "tolerations": {
+            "description": "If specified, the request's tolerations.\n\nTolerations for NoSchedule are required to allocate a device which has a taint with that effect. The same applies to NoExecute.\n\nIn addition, should any of the allocated devices get tainted with NoExecute after allocation and that effect is not tolerated, then all pods consuming the ResourceClaim get deleted to evict them. The scheduler will not let new pods reserve the claim while it has these tainted devices. Once all pods are evicted, the claim will get deallocated.\n\nThe maximum number of tolerations is 16.\n\nThis is an alpha field and requires enabling the DRADeviceTaints feature gate.",
+            "items": {
+              "allOf": [
+                {
+                  "$ref": "#/components/schemas/io.k8s.api.resource.v1beta1.DeviceToleration"
+                }
+              ],
+              "default": {}
+            },
+            "type": "array",
+            "x-kubernetes-list-type": "atomic"
           }
         },
         "required": [
@@ -706,6 +758,66 @@
         ],
         "type": "object"
       },
+      "io.k8s.api.resource.v1beta1.DeviceTaint": {
+        "description": "The device this taint is attached to has the \"effect\" on any claim which does not tolerate the taint and, through the claim, to pods using the claim.",
+        "properties": {
+          "effect": {
+            "default": "",
+            "description": "The effect of the taint on claims that do not tolerate the taint and through such claims on the pods using them. Valid effects are NoSchedule and NoExecute. PreferNoSchedule as used for nodes is not valid here.",
+            "type": "string"
+          },
+          "key": {
+            "default": "",
+            "description": "The taint key to be applied to a device. Must be a label name.",
+            "type": "string"
+          },
+          "timeAdded": {
+            "allOf": [
+              {
+                "$ref": "#/components/schemas/io.k8s.apimachinery.pkg.apis.meta.v1.Time"
+              }
+            ],
+            "description": "TimeAdded represents the time at which the taint was added. Added automatically during create or update if not set."
+          },
+          "value": {
+            "description": "The taint value corresponding to the taint key. Must be a label value.",
+            "type": "string"
+          }
+        },
+        "required": [
+          "key",
+          "effect"
+        ],
+        "type": "object"
+      },
+      "io.k8s.api.resource.v1beta1.DeviceToleration": {
+        "description": "The ResourceClaim this DeviceToleration is attached to tolerates any taint that matches the triple <key,value,effect> using the matching operator <operator>.",
+        "properties": {
+          "effect": {
+            "description": "Effect indicates the taint effect to match. Empty means match all taint effects. When specified, allowed values are NoSchedule and NoExecute.",
+            "type": "string"
+          },
+          "key": {
+            "description": "Key is the taint key that the toleration applies to. Empty means match all taint keys. If the key is empty, operator must be Exists; this combination means to match all values and all keys. Must be a label name.",
+            "type": "string"
+          },
+          "operator": {
+            "default": "Equal",
+            "description": "Operator represents a key's relationship to the value. Valid operators are Exists and Equal. Defaults to Equal. Exists is equivalent to wildcard for value, so that a ResourceClaim can tolerate all taints of a particular category.",
+            "type": "string"
+          },
+          "tolerationSeconds": {
+            "description": "TolerationSeconds represents the period of time the toleration (which must be of effect NoExecute, otherwise this field is ignored) tolerates the taint. By default, it is not set, which means tolerate the taint forever (do not evict). Zero and negative values will be treated as 0 (evict immediately) by the system. If larger than zero, the time when the pod needs to be evicted is calculated as <time when taint was adedd> + <toleration seconds>.",
+            "format": "int64",
+            "type": "integer"
+          },
+          "value": {
+            "description": "Value is the taint value the toleration matches to. If the operator is Exists, the value must be empty, otherwise just a regular string. Must be a label value.",
+            "type": "string"
+          }
+        },
+        "type": "object"
+      },
       "io.k8s.api.resource.v1beta1.NetworkDeviceData": {
         "description": "NetworkDeviceData provides network-related details for the allocated device. This information may be filled by drivers or other components to configure or identify the device within a network context.",
         "properties": {