[GEP-31] Introduce label constants and MachineConfiguration options for in-place updates (#970)

acumino · shafeeqes · ary1992 · web-flow · commit 41fb2af80ebe · 2025-03-06T11:12:11.000+05:30
* Introduce labels constants for InPlaceUpdate

Co-Authored-By: Shafeeque E S &lt;shafeeque.e.s@sap.com&gt;
Co-Authored-By: Ashish Ranjan Yadav &lt;ashish.ranjan.yadav@sap.com&gt;

* Introduce `MachineInPlaceUpdateTimeout` field

Co-authored-by: Shafeeque E S &lt;shafeeque.e.s@sap.com&gt;
Co-authored-by: Ashish Ranjan Yadav  &lt;ashish.ranjan.yadav@sap.com&gt;

* Introduce `DisableHealthTimeout` field

Co-authored-by: Shafeeque E S &lt;shafeeque.e.s@sap.com&gt;
Co-authored-by: Ashish Ranjan Yadav  &lt;ashish.ranjan.yadav@sap.com&gt;

* Run `make generate`

Co-authored-by: Shafeeque E S &lt;shafeeque.e.s@sap.com&gt;
Co-authored-by: Ashish Ranjan Yadav  &lt;ashish.ranjan.yadav@sap.com&gt;

* Nits

* Address Review

* Address Review

---------

Co-authored-by: Shafeeque E S &lt;shafeeque.e.s@sap.com&gt;
Co-authored-by: Ashish Ranjan Yadav &lt;ashish.ranjan.yadav@sap.com&gt;
diff --git a/docs/documents/apis.md b/docs/documents/apis.md
@@ -1055,6 +1055,36 @@ Kubernetes meta/v1.Duration
 </tr>
 <tr>
 <td>
+<code>inPlaceUpdateTimeout</code>
+</td>
+<td>
+<em>
+<a href="https://godoc.org/k8s.io/apimachinery/pkg/apis/meta/v1#Duration">
+Kubernetes meta/v1.Duration
+</a>
+</em>
+</td>
+<td>
+<em>(Optional)</em>
+<p>MachineInPlaceUpdateTimeout is the timeout after which in-place update is declared failed.</p>
+</td>
+</tr>
+<tr>
+<td>
+<code>disableHealthTimeout</code>
+</td>
+<td>
+<em>
+*bool
+</em>
+</td>
+<td>
+<em>(Optional)</em>
+<p>DisableHealthTimeout if set to true, health timeout will be ignored. Leading to machine never being declared failed.</p>
+</td>
+</tr>
+<tr>
+<td>
 <code>maxEvictRetries</code>
 </td>
 <td>
diff --git a/kubernetes/crds/machine.sapcloud.io_machinedeployments.yaml b/kubernetes/crds/machine.sapcloud.io_machinedeployments.yaml
@@ -279,6 +279,11 @@ spec:
                         description: MachineCreationTimeout is the timeout after which
                           machinie creation is declared failed.
                         type: string
+                      disableHealthTimeout:
+                        description: DisableHealthTimeout if set to true, health timeout
+                          will be ignored. Leading to machine never being declared
+                          failed.
+                        type: boolean
                       drainTimeout:
                         description: MachineDraintimeout is the timeout after which
                           machine is forcefully deleted.
@@ -287,6 +292,10 @@ spec:
                         description: MachineHealthTimeout is the timeout after which
                           machine is declared unhealhty/failed.
                         type: string
+                      inPlaceUpdateTimeout:
+                        description: MachineInPlaceUpdateTimeout is the timeout after
+                          which in-place update is declared failed.
+                        type: string
                       maxEvictRetries:
                         description: MaxEvictRetries is the number of retries that
                           will be attempted while draining the node.
diff --git a/kubernetes/crds/machine.sapcloud.io_machines.yaml b/kubernetes/crds/machine.sapcloud.io_machines.yaml
@@ -78,6 +78,10 @@ spec:
                 description: MachineCreationTimeout is the timeout after which machinie
                   creation is declared failed.
                 type: string
+              disableHealthTimeout:
+                description: DisableHealthTimeout if set to true, health timeout will
+                  be ignored. Leading to machine never being declared failed.
+                type: boolean
               drainTimeout:
                 description: MachineDraintimeout is the timeout after which machine
                   is forcefully deleted.
@@ -86,6 +90,10 @@ spec:
                 description: MachineHealthTimeout is the timeout after which machine
                   is declared unhealhty/failed.
                 type: string
+              inPlaceUpdateTimeout:
+                description: MachineInPlaceUpdateTimeout is the timeout after which
+                  in-place update is declared failed.
+                type: string
               maxEvictRetries:
                 description: MaxEvictRetries is the number of retries that will be
                   attempted while draining the node.
diff --git a/kubernetes/crds/machine.sapcloud.io_machinesets.yaml b/kubernetes/crds/machine.sapcloud.io_machinesets.yaml
@@ -161,6 +161,11 @@ spec:
                         description: MachineCreationTimeout is the timeout after which
                           machinie creation is declared failed.
                         type: string
+                      disableHealthTimeout:
+                        description: DisableHealthTimeout if set to true, health timeout
+                          will be ignored. Leading to machine never being declared
+                          failed.
+                        type: boolean
                       drainTimeout:
                         description: MachineDraintimeout is the timeout after which
                           machine is forcefully deleted.
@@ -169,6 +174,10 @@ spec:
                         description: MachineHealthTimeout is the timeout after which
                           machine is declared unhealhty/failed.
                         type: string
+                      inPlaceUpdateTimeout:
+                        description: MachineInPlaceUpdateTimeout is the timeout after
+                          which in-place update is declared failed.
+                        type: string
                       maxEvictRetries:
                         description: MaxEvictRetries is the number of retries that
                           will be attempted while draining the node.
diff --git a/pkg/apis/machine/types.go b/pkg/apis/machine/types.go
@@ -94,6 +94,13 @@ type MachineConfiguration struct {
 	// MachineCreationTimeout is the timeout after which machinie creation is declared failed.
 	MachineCreationTimeout *metav1.Duration
 
+	// MachineInPlaceUpdateTimeout is the timeout after which in-place update is declared failed.
+	MachineInPlaceUpdateTimeout *metav1.Duration
+
+	// DisableHealthTimeout if set to true, health timeout will be ignored. Leading to machine never being declared failed.
+	// This is intended to be used only for in-place updates.
+	DisableHealthTimeout *bool
+
 	// MaxEvictRetries is the number of retries that will be attempted while draining the node.
 	MaxEvictRetries *int32
 
diff --git a/pkg/apis/machine/v1alpha1/constants.go b/pkg/apis/machine/v1alpha1/constants.go
@@ -0,0 +1,22 @@
+// SPDX-FileCopyrightText: 2025 SAP SE or an SAP affiliate company and Gardener contributors
+//
+// SPDX-License-Identifier: Apache-2.0
+
+package v1alpha1
+
+const (
+	// AnnotationKeyMachineUpdateFailedReason is the annotation key that indicates the reason for a machine update failure.
+	AnnotationKeyMachineUpdateFailedReason = "node.machine.sapcloud.io/update-failed-reason"
+
+	// LabelKeyNodeCandidateForUpdate is the label key that indicates a node is a candidate for update.
+	LabelKeyNodeCandidateForUpdate = "node.machine.sapcloud.io/candidate-for-update"
+	// LabelKeyNodeSelectedForUpdate is the label key that indicates a node has been selected for update.
+	LabelKeyNodeSelectedForUpdate = "node.machine.sapcloud.io/selected-for-update"
+	// LabelKeyNodeUpdateResult is the label key that indicates the result of the update on the node.
+	LabelKeyNodeUpdateResult = "node.machine.sapcloud.io/update-result"
+
+	// LabelValueNodeUpdateSuccessful is the label value that indicates the update on the node has succeeded.
+	LabelValueNodeUpdateSuccessful = "successful"
+	// LabelValueNodeUpdateFailed is the label value that indicates the update on the node has failed.
+	LabelValueNodeUpdateFailed = "failed"
+)
diff --git a/pkg/apis/machine/v1alpha1/register.go b/pkg/apis/machine/v1alpha1/register.go
@@ -31,13 +31,6 @@ func Resource(resource string) schema.GroupResource {
 	return SchemeGroupVersion.WithResource(resource).GroupResource()
 }
 
-// func Init() {
-// 	// We only register manually written functions here. The registration of the
-// 	// generated functions takes place in the generated files. The separation
-// 	// makes the code compile even when the generated files are missing.
-// 	SchemeBuilder.Register(addKnownTypes)
-// }
-
 // Adds the list of known types to api.Scheme.
 func addKnownTypes(scheme *runtime.Scheme) error {
 	scheme.AddKnownTypes(SchemeGroupVersion,
diff --git a/pkg/apis/machine/v1alpha1/shared_types.go b/pkg/apis/machine/v1alpha1/shared_types.go
@@ -40,6 +40,15 @@ type MachineConfiguration struct {
 	// +optional
 	MachineCreationTimeout *metav1.Duration `json:"creationTimeout,omitempty"`
 
+	// MachineInPlaceUpdateTimeout is the timeout after which in-place update is declared failed.
+	// +optional
+	MachineInPlaceUpdateTimeout *metav1.Duration `json:"inPlaceUpdateTimeout,omitempty"`
+
+	// DisableHealthTimeout if set to true, health timeout will be ignored. Leading to machine never being declared failed.
+	// This is intended to be used only for in-place updates.
+	// +optional
+	DisableHealthTimeout *bool `json:"disableHealthTimeout,omitempty"`
+
 	// MaxEvictRetries is the number of retries that will be attempted while draining the node.
 	// +optional
 	MaxEvictRetries *int32 `json:"maxEvictRetries,omitempty"`
diff --git a/pkg/apis/machine/v1alpha1/zz_generated.conversion.go b/pkg/apis/machine/v1alpha1/zz_generated.conversion.go
diff --git a/pkg/apis/machine/v1alpha1/zz_generated.deepcopy.go b/pkg/apis/machine/v1alpha1/zz_generated.deepcopy.go
diff --git a/pkg/apis/machine/zz_generated.deepcopy.go b/pkg/apis/machine/zz_generated.deepcopy.go
diff --git a/pkg/controller/machineset.go b/pkg/controller/machineset.go
@@ -339,7 +339,6 @@ func (c *controller) manageReplicas(ctx context.Context, allMachines []*v1alpha1
 	var activeMachines, staleMachines []*v1alpha1.Machine
 	for _, machine := range allMachines {
 		if IsMachineActive(machine) {
-			// klog.Info("Active machine: ", machine.Name)
 			activeMachines = append(activeMachines, machine)
 		} else if IsMachineFailed(machine) {
 			staleMachines = append(staleMachines, machine)
diff --git a/pkg/openapi/api_violations.report b/pkg/openapi/api_violations.report
@@ -5,6 +5,7 @@ API rule violation: list_type_missing,github.com/gardener/machine-controller-man
 API rule violation: names_match,github.com/gardener/machine-controller-manager/pkg/apis/machine/v1alpha1,MachineConfiguration,MachineCreationTimeout
 API rule violation: names_match,github.com/gardener/machine-controller-manager/pkg/apis/machine/v1alpha1,MachineConfiguration,MachineDrainTimeout
 API rule violation: names_match,github.com/gardener/machine-controller-manager/pkg/apis/machine/v1alpha1,MachineConfiguration,MachineHealthTimeout
+API rule violation: names_match,github.com/gardener/machine-controller-manager/pkg/apis/machine/v1alpha1,MachineConfiguration,MachineInPlaceUpdateTimeout
 API rule violation: names_match,github.com/gardener/machine-controller-manager/pkg/apis/machine/v1alpha1,MachineSetStatus,Conditions
 API rule violation: names_match,github.com/gardener/machine-controller-manager/pkg/apis/machine/v1alpha1,MachineSpec,NodeTemplateSpec
 API rule violation: names_match,k8s.io/api/core/v1,AzureDiskVolumeSource,DataDiskURI
diff --git a/pkg/openapi/openapi_generated.go b/pkg/openapi/openapi_generated.go
diff --git a/pkg/util/provider/app/options/options.go b/pkg/util/provider/app/options/options.go
@@ -67,6 +67,7 @@ func NewMCServer() *MCServer {
 				MachineCreationTimeout:                   metav1.Duration{Duration: 20 * time.Minute},
 				MachineHealthTimeout:                     metav1.Duration{Duration: 10 * time.Minute},
 				MachineDrainTimeout:                      metav1.Duration{Duration: drain.DefaultMachineDrainTimeout},
+				MachineInPlaceUpdateTimeout:              metav1.Duration{Duration: 20 * time.Minute},
 				MaxEvictRetries:                          drain.DefaultMaxEvictRetries,
 				PvDetachTimeout:                          metav1.Duration{Duration: 2 * time.Minute},
 				PvReattachTimeout:                        metav1.Duration{Duration: 90 * time.Second},
@@ -100,6 +101,7 @@ func (s *MCServer) AddFlags(fs *pflag.FlagSet) {
 	fs.DurationVar(&s.SafetyOptions.MachineCreationTimeout.Duration, "machine-creation-timeout", s.SafetyOptions.MachineCreationTimeout.Duration, "Timeout (in durartion) used while joining (during creation) of machine before it is declared as failed.")
 	fs.DurationVar(&s.SafetyOptions.MachineHealthTimeout.Duration, "machine-health-timeout", s.SafetyOptions.MachineHealthTimeout.Duration, "Timeout (in durartion) used while re-joining (in case of temporary health issues) of machine before it is declared as failed.")
 	fs.DurationVar(&s.SafetyOptions.MachineDrainTimeout.Duration, "machine-drain-timeout", drain.DefaultMachineDrainTimeout, "Timeout (in durartion) used while draining of machine before deletion, beyond which MCM forcefully deletes machine.")
+	fs.DurationVar(&s.SafetyOptions.MachineInPlaceUpdateTimeout.Duration, "machine-inplace-update-timeout", s.SafetyOptions.MachineInPlaceUpdateTimeout.Duration, "Timeout (in duration) used while updating a machine in-place, beyond which it is declared as failed.")
 	fs.Int32Var(&s.SafetyOptions.MaxEvictRetries, "machine-max-evict-retries", drain.DefaultMaxEvictRetries, "Maximum number of times evicts would be attempted on a pod before it is forcibly deleted during draining of a machine.")
 	fs.DurationVar(&s.SafetyOptions.PvDetachTimeout.Duration, "machine-pv-detach-timeout", s.SafetyOptions.PvDetachTimeout.Duration, "Timeout (in duration) used while waiting for detach of PV while evicting/deleting pods")
 	fs.DurationVar(&s.SafetyOptions.PvReattachTimeout.Duration, "machine-pv-reattach-timeout", s.SafetyOptions.PvReattachTimeout.Duration, "Timeout (in duration) used while waiting for reattach of PV onto a different node")
diff --git a/pkg/util/provider/drain/drain.go b/pkg/util/provider/drain/drain.go
@@ -27,8 +27,6 @@ import (
 	"errors"
 	"fmt"
 	"io"
-	"k8s.io/apimachinery/pkg/labels"
-	"k8s.io/client-go/tools/cache"
 	"regexp"
 	"sort"
 	"strings"
@@ -41,11 +39,13 @@ import (
 	storagev1 "k8s.io/api/storage/v1"
 	apierrors "k8s.io/apimachinery/pkg/api/errors"
 	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+	"k8s.io/apimachinery/pkg/labels"
 	utilerrors "k8s.io/apimachinery/pkg/util/errors"
 	"k8s.io/apimachinery/pkg/util/wait"
 	"k8s.io/client-go/kubernetes"
 	corelisters "k8s.io/client-go/listers/core/v1"
 	policyv1listers "k8s.io/client-go/listers/policy/v1"
+	"k8s.io/client-go/tools/cache"
 	"k8s.io/klog/v2"
 
 	"github.com/gardener/machine-controller-manager/pkg/util/provider/driver"
diff --git a/pkg/util/provider/machineutils/utils.go b/pkg/util/provider/machineutils/utils.go
@@ -6,10 +6,11 @@
 package machineutils
 
 import (
-	"github.com/gardener/machine-controller-manager/pkg/apis/machine/v1alpha1"
 	"time"
 
 	v1 "k8s.io/api/core/v1"
+
+	"github.com/gardener/machine-controller-manager/pkg/apis/machine/v1alpha1"
 )
 
 const (
@@ -69,6 +70,9 @@ const (
 
 	// MachineLabelKey defines the labels which contains the name of the machine of a node
 	MachineLabelKey = "node.gardener.cloud/machine-name"
+
+	// LabelKeyMachineSetScaleUpDisabled is the label key that indicates scaling up of the machine set is disabled.
+	LabelKeyMachineSetScaleUpDisabled = "node.machine.sapcloud.io/scale-up-disabled"
 )
 
 // RetryPeriod is an alias for specifying the retry period
diff --git a/pkg/util/provider/options/types.go b/pkg/util/provider/options/types.go
@@ -87,6 +87,9 @@ type SafetyOptions struct {
 	// Timeout (in duration) used while draining of machine before deletion,
 	// beyond which it forcefully deletes machine
 	MachineDrainTimeout metav1.Duration
+	// Timeout (in duration) used while in-place updating of a machine,
+	// beyond which it is declared as failed
+	MachineInPlaceUpdateTimeout metav1.Duration
 	// Maximum number of times evicts would be attempted on a pod for it is forcibly deleted
 	// during draining of a machine.
 	MaxEvictRetries int32
