mirror of
https://github.com/kubernetes-sigs/descheduler.git
synced 2026-01-26 21:31:18 +01:00
When checking for node limit getting exceeded the pod eviction never fails. Thus, ignoring the metric reporting when a pod fails to be evicted due to node limit constrains. The error also allows plugin to react on other limits getting exceeded. E.g. the limit on the number of pods evicted per namespace.
141 lines
5.0 KiB
Go
141 lines
5.0 KiB
Go
/*
|
|
Copyright 2022 The Kubernetes Authors.
|
|
|
|
Licensed under the Apache License, Version 2.0 (the "License");
|
|
you may not use this file except in compliance with the License.
|
|
You may obtain a copy of the License at
|
|
|
|
http://www.apache.org/licenses/LICENSE-2.0
|
|
|
|
Unless required by applicable law or agreed to in writing, software
|
|
distributed under the License is distributed on an "AS IS" BASIS,
|
|
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
See the License for the specific language governing permissions and
|
|
limitations under the License.
|
|
*/
|
|
|
|
package removepodsviolatingnodetaints
|
|
|
|
import (
|
|
"context"
|
|
"fmt"
|
|
|
|
v1 "k8s.io/api/core/v1"
|
|
"k8s.io/apimachinery/pkg/runtime"
|
|
"k8s.io/apimachinery/pkg/util/sets"
|
|
"k8s.io/klog/v2"
|
|
|
|
"sigs.k8s.io/descheduler/pkg/descheduler/evictions"
|
|
podutil "sigs.k8s.io/descheduler/pkg/descheduler/pod"
|
|
frameworktypes "sigs.k8s.io/descheduler/pkg/framework/types"
|
|
"sigs.k8s.io/descheduler/pkg/utils"
|
|
)
|
|
|
|
const PluginName = "RemovePodsViolatingNodeTaints"
|
|
|
|
// RemovePodsViolatingNodeTaints evicts pods on the node which violate NoSchedule Taints on nodes
|
|
type RemovePodsViolatingNodeTaints struct {
|
|
handle frameworktypes.Handle
|
|
args *RemovePodsViolatingNodeTaintsArgs
|
|
taintFilterFnc func(taint *v1.Taint) bool
|
|
podFilter podutil.FilterFunc
|
|
}
|
|
|
|
var _ frameworktypes.DeschedulePlugin = &RemovePodsViolatingNodeTaints{}
|
|
|
|
// New builds plugin from its arguments while passing a handle
|
|
func New(args runtime.Object, handle frameworktypes.Handle) (frameworktypes.Plugin, error) {
|
|
nodeTaintsArgs, ok := args.(*RemovePodsViolatingNodeTaintsArgs)
|
|
if !ok {
|
|
return nil, fmt.Errorf("want args to be of type RemovePodsViolatingNodeTaintsArgs, got %T", args)
|
|
}
|
|
|
|
var includedNamespaces, excludedNamespaces sets.Set[string]
|
|
if nodeTaintsArgs.Namespaces != nil {
|
|
includedNamespaces = sets.New(nodeTaintsArgs.Namespaces.Include...)
|
|
excludedNamespaces = sets.New(nodeTaintsArgs.Namespaces.Exclude...)
|
|
}
|
|
|
|
// We can combine Filter and PreEvictionFilter since for this strategy it does not matter where we run PreEvictionFilter
|
|
podFilter, err := podutil.NewOptions().
|
|
WithFilter(podutil.WrapFilterFuncs(handle.Evictor().Filter, handle.Evictor().PreEvictionFilter)).
|
|
WithNamespaces(includedNamespaces).
|
|
WithoutNamespaces(excludedNamespaces).
|
|
WithLabelSelector(nodeTaintsArgs.LabelSelector).
|
|
BuildFilterFunc()
|
|
if err != nil {
|
|
return nil, fmt.Errorf("error initializing pod filter function: %v", err)
|
|
}
|
|
|
|
includedTaints := sets.New(nodeTaintsArgs.IncludedTaints...)
|
|
includeTaint := func(taint *v1.Taint) bool {
|
|
// Include only taints by key *or* key=value
|
|
// Always returns true if no includedTaints argument is provided
|
|
return (nodeTaintsArgs.IncludedTaints == nil) || includedTaints.Has(taint.Key) || (taint.Value != "" && includedTaints.Has(fmt.Sprintf("%s=%s", taint.Key, taint.Value)))
|
|
}
|
|
|
|
excludedTaints := sets.New(nodeTaintsArgs.ExcludedTaints...)
|
|
excludeTaint := func(taint *v1.Taint) bool {
|
|
// Exclude taints by key *or* key=value
|
|
return excludedTaints.Has(taint.Key) || (taint.Value != "" && excludedTaints.Has(fmt.Sprintf("%s=%s", taint.Key, taint.Value)))
|
|
}
|
|
|
|
taintFilterFnc := func(taint *v1.Taint) bool {
|
|
return (taint.Effect == v1.TaintEffectNoSchedule) && !excludeTaint(taint) && includeTaint(taint)
|
|
}
|
|
if nodeTaintsArgs.IncludePreferNoSchedule {
|
|
taintFilterFnc = func(taint *v1.Taint) bool {
|
|
return (taint.Effect == v1.TaintEffectNoSchedule || taint.Effect == v1.TaintEffectPreferNoSchedule) && !excludeTaint(taint) && includeTaint(taint)
|
|
}
|
|
}
|
|
|
|
return &RemovePodsViolatingNodeTaints{
|
|
handle: handle,
|
|
podFilter: podFilter,
|
|
args: nodeTaintsArgs,
|
|
taintFilterFnc: taintFilterFnc,
|
|
}, nil
|
|
}
|
|
|
|
// Name retrieves the plugin name
|
|
func (d *RemovePodsViolatingNodeTaints) Name() string {
|
|
return PluginName
|
|
}
|
|
|
|
// Deschedule extension point implementation for the plugin
|
|
func (d *RemovePodsViolatingNodeTaints) Deschedule(ctx context.Context, nodes []*v1.Node) *frameworktypes.Status {
|
|
for _, node := range nodes {
|
|
klog.V(1).InfoS("Processing node", "node", klog.KObj(node))
|
|
pods, err := podutil.ListAllPodsOnANode(node.Name, d.handle.GetPodsAssignedToNodeFunc(), d.podFilter)
|
|
if err != nil {
|
|
// no pods evicted as error encountered retrieving evictable Pods
|
|
return &frameworktypes.Status{
|
|
Err: fmt.Errorf("error listing pods on a node: %v", err),
|
|
}
|
|
}
|
|
totalPods := len(pods)
|
|
loop:
|
|
for i := 0; i < totalPods; i++ {
|
|
if !utils.TolerationsTolerateTaintsWithFilter(
|
|
pods[i].Spec.Tolerations,
|
|
node.Spec.Taints,
|
|
d.taintFilterFnc,
|
|
) {
|
|
klog.V(2).InfoS("Not all taints with NoSchedule effect are tolerated after update for pod on node", "pod", klog.KObj(pods[i]), "node", klog.KObj(node))
|
|
err := d.handle.Evictor().Evict(ctx, pods[i], evictions.EvictOptions{StrategyName: PluginName})
|
|
if err == nil {
|
|
continue
|
|
}
|
|
switch err.(type) {
|
|
case *evictions.EvictionNodeLimitError:
|
|
break loop
|
|
default:
|
|
klog.Errorf("eviction failed: %v", err)
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
return nil
|
|
}
|