autoscaler

README

Scaling documentation has moved to the docs folder.

Documentation

Overview

Package autoscaler calculates the number of pods necessary for the desired level of concurrency per pod (stableConcurrencyPerPod). It operates in two modes, stable mode and panic mode.

Stable mode calculates the average concurrency observed over the last 60 seconds and adjusts the observed pod count to achieve the target value. Current observed pod count is the number of unique pod names which show up in the last 60 seconds.

Panic mode calculates the average concurrency observed over the last 6 seconds and adjusts the observed pod count to achieve the stable target value. Panic mode is engaged when the observed 6 second average concurrency reaches 2x the target stable concurrency. Panic mode will last at least 60 seconds--longer if the 2x threshold is repeatedly breached. During panic mode the number of pods is never decreased in order to prevent flapping.

Index

• Constants
• type Autoscaler
  • func New(config *Config, model v1alpha1.RevisionRequestConcurrencyModelType, ...) *Autoscaler
  • func (a *Autoscaler) Record(ctx context.Context, stat Stat)
  • func (a *Autoscaler) Scale(ctx context.Context, now time.Time) (int32, bool)
• type Config
  • func NewConfigFromConfigMap(configMap *corev1.ConfigMap) (*Config, error)
  • func NewConfigFromMap(data map[string]string) (*Config, error)
  • func (c *Config) TargetConcurrency(model v1alpha1.RevisionRequestConcurrencyModelType) float64
• type Measurement
• type Reporter
  • func NewStatsReporter(podNamespace string, config string, revision string) (*Reporter, error)
  • func (r *Reporter) Report(m Measurement, v float64) error
• type Stat
• type StatMessage
• type StatsReporter

Constants

const (
	ConfigName = "config-autoscaler"
)

Types

type Autoscaler

type Autoscaler struct {
	*Config
	// contains filtered or unexported fields
}

Autoscaler stores current state of an instance of an autoscaler

func New

func New(config *Config, model v1alpha1.RevisionRequestConcurrencyModelType, reporter StatsReporter) *Autoscaler

New creates a new instance of autoscaler

func (*Autoscaler) Record

func (a *Autoscaler) Record(ctx context.Context, stat Stat)

Record a data point.

func (*Autoscaler) Scale

func (a *Autoscaler) Scale(ctx context.Context, now time.Time) (int32, bool)

Scale calculates the desired scale based on current statistics given the current time.

type Config

type Config struct {
	// Feature flags.
	EnableScaleToZero bool
	EnableVPA         bool

	// Target concurrency knobs for different concurrency modes.
	SingleTargetConcurrency   float64
	MultiTargetConcurrency    float64
	VPAMultiTargetConcurrency float64

	// General autoscaler algorithm configuration.
	MaxScaleUpRate           float64
	StableWindow             time.Duration
	PanicWindow              time.Duration
	ScaleToZeroThreshold     time.Duration
	ConcurrencyQuantumOfTime time.Duration
}

Config defines the tunable autoscaler parameters

func NewConfigFromConfigMap

func NewConfigFromConfigMap(configMap *corev1.ConfigMap) (*Config, error)

NewConfigFromConfigMap creates a Config from the supplied ConfigMap

func NewConfigFromMap

func NewConfigFromMap(data map[string]string) (*Config, error)

NewConfigFromMap creates a Config from the supplied map

func (*Config) TargetConcurrency

func (c *Config) TargetConcurrency(model v1alpha1.RevisionRequestConcurrencyModelType) float64

type Measurement

type Measurement int

Measurement represents the type of the autoscaler metric to be reported

const (
	// DesiredPodCountM is used for the pod count that autoscaler wants
	DesiredPodCountM Measurement = iota
	// RequestedPodCountM is used for the requested pod count from kubernetes
	RequestedPodCountM
	// ActualPodCountM is used for the actual number of pods we have
	ActualPodCountM
	// ObservedPodCountM is used for the observed number of pods we have
	ObservedPodCountM
	// ObservedStableConcurrencyM is the average of requests count in each 60 second stable window
	ObservedStableConcurrencyM
	// ObservedPanicConcurrencyM is the average of requests count in each 6 second panic window
	ObservedPanicConcurrencyM
	// TargetConcurrencyM is the desired number of concurrent requests for each pod
	TargetConcurrencyM
	// PanicM is used as a flag to indicate if autoscaler is in panic mode or not
	PanicM
)

type Reporter

type Reporter struct {
	// contains filtered or unexported fields
}

Reporter holds cached metric objects to report autoscaler metrics

func NewStatsReporter

func NewStatsReporter(podNamespace string, config string, revision string) (*Reporter, error)

NewStatsReporter creates a reporter that collects and reports autoscaler metrics

func (*Reporter) Report

func (r *Reporter) Report(m Measurement, v float64) error

Report captures value v for measurement m

type Stat

type Stat struct {
	// The time the data point was collected on the pod.
	Time *time.Time

	// The unique identity of this pod.  Used to count how many pods
	// are contributing to the metrics.
	PodName string

	// Average number of requests currently being handled by this pod.
	AverageConcurrentRequests float64

	// Number of requests received since last Stat (approximately QPS).
	RequestCount int32
}

Stat defines a single measurement at a point in time

type StatMessage

type StatMessage struct {
	RevisionKey string
	Stat        Stat
}

StatMessage wraps a Stat with identifying information so it can be routed to the correct receiver.

type StatsReporter

type StatsReporter interface {
	Report(m Measurement, v float64) error
}

StatsReporter defines the interface for sending autoscaler metrics