k8s教程说明

prometheus全组件的教程

go语言课程

告警的ql


histogram_quantile(0.99, sum(rate(scheduler_e2e_scheduling_duration_seconds_bucket{job="kube-scheduler"}[5m])) without(instance, pod)) > 3 for 1m

含义：调度耗时超过3秒
追踪这个 histogram的metrics
代码版本 v1.20
位置 D:\go_path\src\github.com\kubernetes\kubernetes\pkg\scheduler\metrics\metrics.go
追踪调用方，在observeScheduleAttemptAndLatency的封装中，位置 D:\go_path\src\github.com\kubernetes\kubernetes\pkg\scheduler\metrics\profile_metrics.go
这里可看到调度的三种结果都会记录相关的耗时

追踪调用方

位置 D:\go_path\src\github.com\kubernetes\kubernetes\pkg\scheduler\scheduler.go + 608
在函数 Scheduler.scheduleOne中，用来记录调度每个pod的耗时
可以看到具体的调用点，在异步bind函数的底部

由此得出结论 e2e 是统计整个scheduleOne的耗时

go func() {
      err := sched.bind(bindingCycleCtx, fwk, assumedPod, scheduleResult.SuggestedHost, state)
      if err != nil {
          metrics.PodScheduleError(fwk.ProfileName(), metrics.SinceInSeconds(start))
          // trigger un-reserve plugins to clean up state associated with the reserved Pod
          fwk.RunReservePluginsUnreserve(bindingCycleCtx, state, assumedPod, scheduleResult.SuggestedHost)
          if err := sched.SchedulerCache.ForgetPod(assumedPod); err != nil {
              klog.Errorf("scheduler cache ForgetPod failed: %v", err)
          }
          sched.recordSchedulingFailure(fwk, assumedPodInfo, fmt.Errorf("binding rejected: %w", err), SchedulerError, "")
      } else {
          // Calculating nodeResourceString can be heavy. Avoid it if klog verbosity is below 2.
          if klog.V(2).Enabled() {
              klog.InfoS("Successfully bound pod to node", "pod", klog.KObj(pod), "node", scheduleResult.SuggestedHost, "evaluatedNodes", scheduleResult.EvaluatedNodes, "feasibleNodes", scheduleResult.FeasibleNodes)
          }
          metrics.PodScheduled(fwk.ProfileName(), metrics.SinceInSeconds(start))
          metrics.PodSchedulingAttempts.Observe(float64(podInfo.Attempts))
          metrics.PodSchedulingDuration.WithLabelValues(getAttemptsLabel(podInfo)).Observe(metrics.SinceInSeconds(podInfo.InitialAttemptTimestamp))

          // Run "postbind" plugins.
          fwk.RunPostBindPlugins(bindingCycleCtx, state, assumedPod, scheduleResult.SuggestedHost)
      }
}

scheduleOne从上到下都包含哪几个过程

01 调度算法耗时

实例代码

// 调用调度算法给出结果
scheduleResult, err := sched.Algorithm.Schedule(schedulingCycleCtx, fwk, state, pod)
// 处理错误
if err != nil{}
// 记录调度算法耗时
metrics.SchedulingAlgorithmLatency.Observe(metrics.SinceInSeconds(start
}))

从上面可以看出主要分3个步骤
- 调用调度算法给出结果
- 处理错误
- 记录调度算法耗时

那么我们首先应该算法的耗时，对应的histogram metrics为

histogram_quantile(0.99, sum(rate(scheduler_scheduling_algorithm_duration_seconds_bucket{job="kube-scheduler"}[5m])) by (le))

将e2e和algorithm 99分位耗时再结合告警时间的曲线发现吻合度较高
但是发现99分位下 algorithm > e2e ，但是按照e2e作为兜底来看，应该是e2e要更高，所以调整999分位发现2者差不多
造成上述问题的原因跟prometheus histogram线性插值法的误差有关系，具体可以看我的文章 histogram线性插值法原理

Algorithm.Schedule具体流程

在Schedule中可以看到两个主要的函数调用


feasibleNodes, filteredNodesStatuses, err := g.findNodesThatFitPod(ctx, fwk, state, pod)
priorityList, err := g.prioritizeNodes(ctx, fwk, state, pod, feasibleNodes)

其中 findNodesThatFitPod 对应的是filter流程，对应的metrics有 scheduler_framework_extension_point_duration_seconds_bucket

histogram_quantile(0.999, sum by(extension_point,le) (rate(scheduler_framework_extension_point_duration_seconds_bucket{job="kube-scheduler"}[5m])))

相关的截图可以看到

prioritizeNodes对应的是score流程，对应的metrics有

histogram_quantile(0.99, sum by(plugin,le) (rate(scheduler_plugin_execution_duration_seconds_bucket{job="kube-scheduler"}[5m])))

相关的截图可以看到
上述具体的算法流程可以和官方文档给出的流程图对得上

02 调度算法耗时

再回过头来看bind的过程

其中的核心就在bind这里

err := sched.bind(bindingCycleCtx, fwk, assumedPod, scheduleResult.SuggestedHost, state)

可以看到在bind函数内部是单独计时的

func (sched *Scheduler) bind(ctx context.Context, fwk framework.Framework, assumed *v1.Pod, targetNode string, state *framework.CycleState) (err error) {
  start := time.Now()
  defer func() {
      sched.finishBinding(fwk, assumed, targetNode, start, err)
  }()

  bound, err := sched.extendersBinding(assumed, targetNode)
  if bound {
      return err
  }
  bindStatus := fwk.RunBindPlugins(ctx, state, assumed, targetNode)
  if bindStatus.IsSuccess() {
      return nil
  }
  if bindStatus.Code() == framework.Error {
      return bindStatus.AsError()
  }
  return fmt.Errorf("bind status: %s, %v", bindStatus.Code().String(), bindStatus.Message())
}

对应的metric为

histogram_quantile(0.999, sum by(le) (rate(scheduler_binding_duration_seconds_bucket{job="kube-scheduler"}[5m])))

这里我们对比e2e和bind的999分位值
发现相比于alg，bind和e2e吻合度更高
同时发现bind内部主要两个流程 sched.extendersBinding执行外部binding插件
fwk.RunBindPlugins 执行内部的绑定插件

内部绑定插件

代码如下，主要流程就是执行绑定插件

// RunBindPlugins runs the set of configured bind plugins until one returns a non `Skip` status.
func (f *frameworkImpl) RunBindPlugins(ctx context.Context, state *framework.CycleState, pod *v1.Pod, nodeName string) (status *framework.Status) {
  startTime := time.Now()
  defer func() {
      metrics.FrameworkExtensionPointDuration.WithLabelValues(bind, status.Code().String(), f.profileName).Observe(metrics.SinceInSeconds(startTime))
  }()
  if len(f.bindPlugins) == 0 {
      return framework.NewStatus(framework.Skip, "")
  }
  for _, bp := range f.bindPlugins {
      status = f.runBindPlugin(ctx, bp, state, pod, nodeName)
      if status != nil && status.Code() == framework.Skip {
          continue
      }
      if !status.IsSuccess() {
          err := status.AsError()
          klog.ErrorS(err, "Failed running Bind plugin", "plugin", bp.Name(), "pod", klog.KObj(pod))
          return framework.AsStatus(fmt.Errorf("running Bind plugin %q: %w", bp.Name(), err))
      }
      return status
  }
  return status
}

那么默认的绑定插件为调用 pod的bind方法绑定到指定的node上，binding是pods的子资源

// Bind binds pods to nodes using the k8s client.
func (b DefaultBinder) Bind(ctx context.Context, state *framework.CycleState, p *v1.Pod, nodeName string) *framework.Status {
  klog.V(3).Infof("Attempting to bind %v/%v to %v", p.Namespace, p.Name, nodeName)
  binding := &v1.Binding{
      ObjectMeta: metav1.ObjectMeta{Namespace: p.Namespace, Name: p.Name, UID: p.UID},
      Target:     v1.ObjectReference{Kind: "Node", Name: nodeName},
  }
  err := b.handle.ClientSet().CoreV1().Pods(binding.Namespace).Bind(ctx, binding, metav1.CreateOptions{})
  if err != nil {
      return framework.AsStatus(err)
  }
  return nil
}

执行绑定动作也有相关的metrics统计耗时，

histogram_quantile(0.999, sum by(le) (rate(scheduler_plugin_execution_duration_seconds_bucket{extension_point="Bind",plugin="DefaultBinder",job="kube-scheduler"}[5m])))

同时在 RunBindPlugins中也有defer func负责统计耗时

histogram_quantile(0.9999, sum by(le) (rate(scheduler_framework_extension_point_duration_seconds_bucket{extension_point="Bind",job="kube-scheduler"}[5m])))

从上面两个metrics看，内部的插件耗时都很低

extendersBinding 外部插件

代码如下，遍历Algorithm的Extenders，判断是bind类型的，然后执行extender.Bind

// TODO(#87159): Move this to a Plugin.
func (sched *Scheduler) extendersBinding(pod *v1.Pod, node string) (bool, error) {
  for _, extender := range sched.Algorithm.Extenders() {
      if !extender.IsBinder() || !extender.IsInterested(pod) {
          continue
      }
      return true, extender.Bind(&v1.Binding{
          ObjectMeta: metav1.ObjectMeta{Namespace: pod.Namespace, Name: pod.Name, UID: pod.UID},
          Target:     v1.ObjectReference{Kind: "Node", Name: node},
      })
  }
  return false, nil
}

extender.Bind对应就是通过http发往外部的调度器

// Bind delegates the action of binding a pod to a node to the extender.
func (h *HTTPExtender) Bind(binding *v1.Binding) error {
  var result extenderv1.ExtenderBindingResult
  if !h.IsBinder() {
      // This shouldn't happen as this extender wouldn't have become a Binder.
      return fmt.Errorf("unexpected empty bindVerb in extender")
  }
  req := &extenderv1.ExtenderBindingArgs{
      PodName:      binding.Name,
      PodNamespace: binding.Namespace,
      PodUID:       binding.UID,
      Node:         binding.Target.Name,
  }
  if err := h.send(h.bindVerb, req, &result); err != nil {
      return err
  }
  if result.Error != "" {
      return fmt.Errorf(result.Error)
  }
  return nil
}

很遗憾的是这里并没有相关的metrics统计耗时
目前猜测遍历 sched.Algorithm.Extenders 执行的耗时
这里sched.Algorithm.Extenders来自于 KubeSchedulerConfiguration 中的配置
也就是编写配置文件，并将其路径传给 kube-scheduler 的命令行参数，定制 kube-scheduler 的行为，目前并没有看到

总结

scheduler 调度过程

单个pod的调度主要分为3个步骤：
- 根据Predict和Priority两个阶段，调用各自的算法插件，选择最优的Node
- Assume这个Pod被调度到对应的Node，保存到cache
- 用extender和plugins进行验证，如果通过则绑定
e2e 耗时主要来自bind
但目前看到bind执行耗时并没有很长
待续

从k8s集群e2e调度慢告警看kube-scheduler源码