kubernetes v1.18减少了HPA v2beta2的behavior字段，能够精细化的管制伸缩的行为。

若不指定behavior字段，则按默认的behavior行为执行伸缩。

一. 默认behavior

默认的behavior:

behavior:  scaleDown:    stabilizationWindowSeconds: 300            // 冷却工夫=5min    policies:    - type: Percent      value: 100      periodSeconds: 15  scaleUp:    stabilizationWindowSeconds: 0            // 没有冷却工夫    policies:    - type: Percent      value: 100      periodSeconds: 15                        // replicas翻倍扩容    - type: Pods      value: 4      periodSeconds: 15                        // replicas扩容4个    selectPolicy: Max                        // 取最大的策略

扩容时，疾速扩容：

  • 不思考历史计算值(stabilizationWindowSeconds=0)，即reconcile()发现指标变动时，计算新replicas后，立刻执行伸缩；

  • 每15秒最多容许：

    • 正本翻倍(减少currentReplicas*100%个正本) 或者 每15s新增4个正本，两者取max；
    • 即：max(2*currentReplicas, 4)

缩容时，迟缓缩容：

  • 缩容后的最终正本数不得<过来5min的历史最大值(stabilizationWindowSeconds=300)，即冷却工夫=300s；
  • 每15秒最多容许：正本缩小currentReplicas*100%个正本；

二. demo

扩容，将指标猛增(1-->13)：

  • 首先，从1正本扩容到4正本；
  • 而后，从4正本扩容到8正本；
  • 最初，从8正本扩容至指标计算的13正本；
# kubectl describe hpaName:                    sample-appNamespace:               defaultLabels:                  <none>Annotations:             <none>Reference:               Deployment/sample-appMetrics:                 ( current / target )  "metric_hpa" on pods:  1 / 1Min replicas:            1Max replicas:            15Deployment pods:         13 current / 13 desiredConditions:  Type            Status  Reason              Message  ----            ------  ------              -------  AbleToScale     True    ReadyForNewScale    recommended size matches current size  ScalingActive   True    ValidMetricFound    the HPA was able to successfully calculate a replica count from pods metric metric_hpa  ScalingLimited  False   DesiredWithinRange  the desired count is within the acceptable rangeEvents:  Type    Reason             Age    From                       Message  ----    ------             ----   ----                       -------  Normal  SuccessfulRescale  2m14s  horizontal-pod-autoscaler  New size: 4; reason: pods metric metric_hpa above target  Normal  SuccessfulRescale  119s   horizontal-pod-autoscaler  New size: 8; reason: pods metric metric_hpa above target  Normal  SuccessfulRescale  103s   horizontal-pod-autoscaler  New size: 13; reason: pods metric metric_hpa above target

缩容，将指标猛降(13-->1):

  • 5min后，hpa间接将replicas缩容至minReplicas=1；
# kubectl describe hpaName:                    sample-appNamespace:               defaultLabels:                  <none>Annotations:             <none>Reference:               Deployment/sample-appMetrics:                 ( current / target )  "metric_hpa" on pods:  1 / 1Min replicas:            1Max replicas:            15Deployment pods:         1 current / 1 desiredConditions:  Type            Status  Reason              Message  ----            ------  ------              -------  AbleToScale     True    ReadyForNewScale    recommended size matches current size  ScalingActive   True    ValidMetricFound    the HPA was able to successfully calculate a replica count from pods metric metric_hpa  ScalingLimited  False   DesiredWithinRange  the desired count is within the acceptable rangeEvents:  Type    Reason             Age   From                       Message  ----    ------             ----  ----                       -------  Normal  SuccessfulRescale  11m   horizontal-pod-autoscaler  New size: 4; reason: pods metric metric_hpa above target  Normal  SuccessfulRescale  11m   horizontal-pod-autoscaler  New size: 8; reason: pods metric metric_hpa above target  Normal  SuccessfulRescale  11m   horizontal-pod-autoscaler  New size: 13; reason: pods metric metric_hpa above target  Normal  SuccessfulRescale  98s   horizontal-pod-autoscaler  New size: 1; reason: All metrics below target

三. 源码剖析

基于v1.20源码。

代码入口：

  • 输出参数：prenormalizedDesiredReplicas = 依据指标计算的所需replicas；
  • 首先，依据默认的冷却工夫，计算stabilizedRecommendation正本数；
  • 而后，依据stabilizedRecommendation正本数 + 默认的伸缩策略，计算最终的正本数；
// pkg/controller/podautoscaler/horizontal.gofunc (a *HorizontalController) normalizeDesiredReplicas(hpa *autoscalingv2.HorizontalPodAutoscaler, key string, currentReplicas int32, prenormalizedDesiredReplicas int32, minReplicas int32) int32 {    // 冷却工夫    stabilizedRecommendation := a.stabilizeRecommendation(key, prenormalizedDesiredReplicas)    ...    // 伸缩策略    desiredReplicas, condition, reason := convertDesiredReplicasWithRules(currentReplicas, stabilizedRecommendation, minReplicas, hpa.Spec.MaxReplicas)    ...    return desiredReplicas}

先看下冷却工夫：

  • 计算maxRecommendation的值：

    • 初始=prenormalizedDesiredReplicas，即指标计算的replicas；

  • 遍历历史伸缩记录：

    • 若5min内有伸缩 && 伸缩replicas > maxRecommendation，则笼罩maxRecommendation；

  • 根据上述逻辑：

    • 扩容时，最近的oldReplicas < newReplicas，故最终返回值=prenormalizedDesiredReplicas，即立刻扩容；
    • 缩容时，要抉择max(5min内oldReplicas，newReplicas)，因为缩容时oldReplicas > newReplicas，故要期待5min后能力执行缩容；
// pkg/controller/podautoscaler/horizontal.gofunc (a *HorizontalController) stabilizeRecommendation(key string, prenormalizedDesiredReplicas int32) int32 {    maxRecommendation := prenormalizedDesiredReplicas    foundOldSample := false    oldSampleIndex := 0    cutoff := time.Now().Add(-a.downscaleStabilisationWindow)    // 300s    for i, rec := range a.recommendations[key] {        if rec.timestamp.Before(cutoff) {            foundOldSample = true            oldSampleIndex = i        } else if rec.recommendation > maxRecommendation {            maxRecommendation = rec.recommendation        }    }    if foundOldSample {        a.recommendations[key][oldSampleIndex] = timestampedRecommendation{prenormalizedDesiredReplicas, time.Now()}    } else {        a.recommendations[key] = append(a.recommendations[key], timestampedRecommendation{prenormalizedDesiredReplicas, time.Now()})    }    return maxRecommendation}

再看下伸缩策略：

  • 扩容时，最终正本数=min(hpaMaxReplicas, max(2*currentReplicas, 4), desiredReplicas)；

    • 也就是说，若计算的desiredReplicas太大的话，会应用max(2*currentReplicas, 4)进行限度；
  • 缩容时，最终正本数=max(desiredReplicas, hpaMinReplicas)；
// pkg/controller/podautoscaler/horizontal.gofunc convertDesiredReplicasWithRules(currentReplicas, desiredReplicas, hpaMinReplicas, hpaMaxReplicas int32) (int32, string, string) {    var minimumAllowedReplicas int32    var maximumAllowedReplicas int32    var possibleLimitingCondition string    var possibleLimitingReason string    minimumAllowedReplicas = hpaMinReplicas    // Do not upscale too much to prevent incorrect rapid increase of the number of master replicas caused by    // bogus CPU usage report from heapster/kubelet (like in issue #32304).    scaleUpLimit := calculateScaleUpLimit(currentReplicas)    if hpaMaxReplicas > scaleUpLimit {        maximumAllowedReplicas = scaleUpLimit        possibleLimitingCondition = "ScaleUpLimit"        possibleLimitingReason = "the desired replica count is increasing faster than the maximum scale rate"    } else {        maximumAllowedReplicas = hpaMaxReplicas        possibleLimitingCondition = "TooManyReplicas"        possibleLimitingReason = "the desired replica count is more than the maximum replica count"    }    if desiredReplicas < minimumAllowedReplicas {        possibleLimitingCondition = "TooFewReplicas"        possibleLimitingReason = "the desired replica count is less than the minimum replica count"        return minimumAllowedReplicas, possibleLimitingCondition, possibleLimitingReason    } else if desiredReplicas > maximumAllowedReplicas {        return maximumAllowedReplicas, possibleLimitingCondition, possibleLimitingReason    }    return desiredReplicas, "DesiredWithinRange", "the desired count is within the acceptable range"}
func calculateScaleUpLimit(currentReplicas int32) int32 {    return int32(math.Max(scaleUpLimitFactor*float64(currentReplicas), scaleUpLimitMinimum))    // scaleUpLimitFactor=2, scaleUpLimitMinimum=4}

参考：

1.https://zhuanlan.zhihu.com/p/245208287