mirror of
https://github.com/ceph/ceph-csi.git
synced 2024-12-04 20:20:19 +00:00
252 lines
9.2 KiB
Go
252 lines
9.2 KiB
Go
/*
|
|
Copyright 2014 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 podutils
|
|
|
|
import (
|
|
"time"
|
|
|
|
corev1 "k8s.io/api/core/v1"
|
|
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
|
|
)
|
|
|
|
// IsPodAvailable returns true if a pod is available; false otherwise.
|
|
// Precondition for an available pod is that it must be ready. On top
|
|
// of that, there are two cases when a pod can be considered available:
|
|
// 1. minReadySeconds == 0, or
|
|
// 2. LastTransitionTime (is set) + minReadySeconds < current time
|
|
func IsPodAvailable(pod *corev1.Pod, minReadySeconds int32, now metav1.Time) bool {
|
|
if !IsPodReady(pod) {
|
|
return false
|
|
}
|
|
|
|
c := getPodReadyCondition(pod.Status)
|
|
minReadySecondsDuration := time.Duration(minReadySeconds) * time.Second
|
|
if minReadySeconds == 0 || !c.LastTransitionTime.IsZero() && c.LastTransitionTime.Add(minReadySecondsDuration).Before(now.Time) {
|
|
return true
|
|
}
|
|
return false
|
|
}
|
|
|
|
// IsPodReady returns true if a pod is ready; false otherwise.
|
|
func IsPodReady(pod *corev1.Pod) bool {
|
|
return isPodReadyConditionTrue(pod.Status)
|
|
}
|
|
|
|
func isPodDeleting(pod *corev1.Pod) bool {
|
|
return pod.DeletionTimestamp != nil
|
|
}
|
|
|
|
// IsPodReadyConditionTrue returns true if a pod is ready; false otherwise.
|
|
func isPodReadyConditionTrue(status corev1.PodStatus) bool {
|
|
condition := getPodReadyCondition(status)
|
|
return condition != nil && condition.Status == corev1.ConditionTrue
|
|
}
|
|
|
|
// GetPodReadyCondition extracts the pod ready condition from the given status and returns that.
|
|
// Returns nil if the condition is not present.
|
|
func getPodReadyCondition(status corev1.PodStatus) *corev1.PodCondition {
|
|
_, condition := getPodCondition(&status, corev1.PodReady)
|
|
return condition
|
|
}
|
|
|
|
// GetPodCondition extracts the provided condition from the given status and returns that.
|
|
// Returns nil and -1 if the condition is not present, and the index of the located condition.
|
|
func getPodCondition(status *corev1.PodStatus, conditionType corev1.PodConditionType) (int, *corev1.PodCondition) {
|
|
if status == nil {
|
|
return -1, nil
|
|
}
|
|
return getPodConditionFromList(status.Conditions, conditionType)
|
|
}
|
|
|
|
// GetPodConditionFromList extracts the provided condition from the given list of condition and
|
|
// returns the index of the condition and the condition. Returns -1 and nil if the condition is not present.
|
|
func getPodConditionFromList(conditions []corev1.PodCondition, conditionType corev1.PodConditionType) (int, *corev1.PodCondition) {
|
|
if conditions == nil {
|
|
return -1, nil
|
|
}
|
|
for i := range conditions {
|
|
if conditions[i].Type == conditionType {
|
|
return i, &conditions[i]
|
|
}
|
|
}
|
|
return -1, nil
|
|
}
|
|
|
|
// ByLogging allows custom sorting of pods so the best one can be picked for getting its logs.
|
|
type ByLogging []*corev1.Pod
|
|
|
|
func (s ByLogging) Len() int { return len(s) }
|
|
func (s ByLogging) Swap(i, j int) { s[i], s[j] = s[j], s[i] }
|
|
|
|
func (s ByLogging) Less(i, j int) bool {
|
|
// 1. assigned < unassigned
|
|
if s[i].Spec.NodeName != s[j].Spec.NodeName && (len(s[i].Spec.NodeName) == 0 || len(s[j].Spec.NodeName) == 0) {
|
|
return len(s[i].Spec.NodeName) > 0
|
|
}
|
|
// 2. PodRunning < PodUnknown < PodPending
|
|
m := map[corev1.PodPhase]int{corev1.PodRunning: 0, corev1.PodUnknown: 1, corev1.PodPending: 2}
|
|
if m[s[i].Status.Phase] != m[s[j].Status.Phase] {
|
|
return m[s[i].Status.Phase] < m[s[j].Status.Phase]
|
|
}
|
|
// 3. ready < not ready
|
|
if IsPodReady(s[i]) != IsPodReady(s[j]) {
|
|
return IsPodReady(s[i])
|
|
}
|
|
// TODO: take availability into account when we push minReadySeconds information from deployment into pods,
|
|
// see https://github.com/kubernetes/kubernetes/issues/22065
|
|
// 4. Been ready for more time < less time < empty time
|
|
if IsPodReady(s[i]) && IsPodReady(s[j]) && !podReadyTime(s[i]).Equal(podReadyTime(s[j])) {
|
|
return afterOrZero(podReadyTime(s[j]), podReadyTime(s[i]))
|
|
}
|
|
// 5. Pods with containers with higher restart counts < lower restart counts
|
|
if maxContainerRestarts(s[i]) != maxContainerRestarts(s[j]) {
|
|
return maxContainerRestarts(s[i]) > maxContainerRestarts(s[j])
|
|
}
|
|
// 6. older pods < newer pods < empty timestamp pods
|
|
if !s[i].CreationTimestamp.Equal(&s[j].CreationTimestamp) {
|
|
return afterOrZero(&s[j].CreationTimestamp, &s[i].CreationTimestamp)
|
|
}
|
|
return false
|
|
}
|
|
|
|
// ActivePods type allows custom sorting of pods so a controller can pick the best ones to delete.
|
|
type ActivePods []*corev1.Pod
|
|
|
|
func (s ActivePods) Len() int { return len(s) }
|
|
func (s ActivePods) Swap(i, j int) { s[i], s[j] = s[j], s[i] }
|
|
|
|
func (s ActivePods) Less(i, j int) bool {
|
|
// 1. Unassigned < assigned
|
|
// If only one of the pods is unassigned, the unassigned one is smaller
|
|
if s[i].Spec.NodeName != s[j].Spec.NodeName && (len(s[i].Spec.NodeName) == 0 || len(s[j].Spec.NodeName) == 0) {
|
|
return len(s[i].Spec.NodeName) == 0
|
|
}
|
|
// 2. PodPending < PodUnknown < PodRunning
|
|
m := map[corev1.PodPhase]int{corev1.PodPending: 0, corev1.PodUnknown: 1, corev1.PodRunning: 2}
|
|
if m[s[i].Status.Phase] != m[s[j].Status.Phase] {
|
|
return m[s[i].Status.Phase] < m[s[j].Status.Phase]
|
|
}
|
|
// 3. Not ready < ready
|
|
// If only one of the pods is not ready, the not ready one is smaller
|
|
if IsPodReady(s[i]) != IsPodReady(s[j]) {
|
|
return !IsPodReady(s[i])
|
|
}
|
|
// 4. Deleting < Not deleting
|
|
if isPodDeleting(s[i]) != isPodDeleting(s[j]) {
|
|
return isPodDeleting(s[i])
|
|
}
|
|
// 5. Older deletion timestamp < newer deletion timestamp
|
|
if isPodDeleting(s[i]) && isPodDeleting(s[j]) && !s[i].ObjectMeta.DeletionTimestamp.Equal(s[j].ObjectMeta.DeletionTimestamp) {
|
|
return s[i].ObjectMeta.DeletionTimestamp.Before(s[j].ObjectMeta.DeletionTimestamp)
|
|
}
|
|
// TODO: take availability into account when we push minReadySeconds information from deployment into pods,
|
|
// see https://github.com/kubernetes/kubernetes/issues/22065
|
|
// 6. Been ready for empty time < less time < more time
|
|
// If both pods are ready, the latest ready one is smaller
|
|
if IsPodReady(s[i]) && IsPodReady(s[j]) && !podReadyTime(s[i]).Equal(podReadyTime(s[j])) {
|
|
return afterOrZero(podReadyTime(s[i]), podReadyTime(s[j]))
|
|
}
|
|
// 7. Pods with containers with higher restart counts < lower restart counts
|
|
if maxContainerRestarts(s[i]) != maxContainerRestarts(s[j]) {
|
|
return maxContainerRestarts(s[i]) > maxContainerRestarts(s[j])
|
|
}
|
|
// 8. Empty creation time pods < newer pods < older pods
|
|
if !s[i].CreationTimestamp.Equal(&s[j].CreationTimestamp) {
|
|
return afterOrZero(&s[i].CreationTimestamp, &s[j].CreationTimestamp)
|
|
}
|
|
return false
|
|
}
|
|
|
|
// afterOrZero checks if time t1 is after time t2; if one of them
|
|
// is zero, the zero time is seen as after non-zero time.
|
|
func afterOrZero(t1, t2 *metav1.Time) bool {
|
|
if t1.Time.IsZero() || t2.Time.IsZero() {
|
|
return t1.Time.IsZero()
|
|
}
|
|
return t1.After(t2.Time)
|
|
}
|
|
|
|
func podReadyTime(pod *corev1.Pod) *metav1.Time {
|
|
for _, c := range pod.Status.Conditions {
|
|
// we only care about pod ready conditions
|
|
if c.Type == corev1.PodReady && c.Status == corev1.ConditionTrue {
|
|
return &c.LastTransitionTime
|
|
}
|
|
}
|
|
return &metav1.Time{}
|
|
}
|
|
|
|
func maxContainerRestarts(pod *corev1.Pod) int {
|
|
maxRestarts := 0
|
|
for _, c := range pod.Status.ContainerStatuses {
|
|
maxRestarts = max(maxRestarts, int(c.RestartCount))
|
|
}
|
|
return maxRestarts
|
|
}
|
|
|
|
// ContainerType and VisitContainers are taken from
|
|
// https://github.com/kubernetes/kubernetes/blob/master/pkg/api/v1/pod/util.go
|
|
// kubectl cannot directly import this due to project goals
|
|
|
|
// ContainerType signifies container type
|
|
type ContainerType int
|
|
|
|
const (
|
|
// Containers is for normal containers
|
|
Containers ContainerType = 1 << iota
|
|
// InitContainers is for init containers
|
|
InitContainers
|
|
// EphemeralContainers is for ephemeral containers
|
|
EphemeralContainers
|
|
)
|
|
|
|
// AllContainers specifies that all containers be visited.
|
|
const AllContainers ContainerType = (InitContainers | Containers | EphemeralContainers)
|
|
|
|
// ContainerVisitor is called with each container spec, and returns true
|
|
// if visiting should continue.
|
|
type ContainerVisitor func(container *corev1.Container, containerType ContainerType) (shouldContinue bool)
|
|
|
|
// VisitContainers invokes the visitor function with a pointer to every container
|
|
// spec in the given pod spec with type set in mask. If visitor returns false,
|
|
// visiting is short-circuited. VisitContainers returns true if visiting completes,
|
|
// false if visiting was short-circuited.
|
|
func VisitContainers(podSpec *corev1.PodSpec, mask ContainerType, visitor ContainerVisitor) bool {
|
|
if mask&InitContainers != 0 {
|
|
for i := range podSpec.InitContainers {
|
|
if !visitor(&podSpec.InitContainers[i], InitContainers) {
|
|
return false
|
|
}
|
|
}
|
|
}
|
|
if mask&Containers != 0 {
|
|
for i := range podSpec.Containers {
|
|
if !visitor(&podSpec.Containers[i], Containers) {
|
|
return false
|
|
}
|
|
}
|
|
}
|
|
if mask&EphemeralContainers != 0 {
|
|
for i := range podSpec.EphemeralContainers {
|
|
if !visitor((*corev1.Container)(&podSpec.EphemeralContainers[i].EphemeralContainerCommon), EphemeralContainers) {
|
|
return false
|
|
}
|
|
}
|
|
}
|
|
return true
|
|
}
|