metrics

Documentation

Index

• Variables
• type Chore
  • func NewChore(log *zap.Logger, config Config, loop *segmentloop.Service) *Chore
  • func (chore *Chore) Close() error
  • func (chore *Chore) Run(ctx context.Context) (err error)
• type Config
• type Counter
  • func NewCounter() *Counter
  • func (counter *Counter) InlineSegment(ctx context.Context, segment *segmentloop.Segment) error
  • func (counter *Counter) LoopStarted(context.Context, segmentloop.LoopInfo) (err error)
  • func (counter *Counter) RemoteSegment(ctx context.Context, segment *segmentloop.Segment) error
• type Metrics
  • func (metrics *Metrics) Aggregate(partial Metrics)
  • func (metrics *Metrics) Reset()
• type Observer
  • func NewObserver() *Observer
  • func (obs *Observer) Finish(ctx context.Context) error
  • func (obs *Observer) Fork(ctx context.Context) (rangedloop.Partial, error)
  • func (obs *Observer) Join(ctx context.Context, partial rangedloop.Partial) error
  • func (obs *Observer) Start(ctx context.Context, startTime time.Time) error
  • func (obs *Observer) TestingMetrics() Metrics

Variables

var (
	// Error defines the metrics chore errors class.
	Error = errs.Class("metrics")
)

Types

type Chore

type Chore struct {
	Loop *sync2.Cycle

	Counter *Counter
	// contains filtered or unexported fields
}

Chore implements the metrics chore.

architecture: Chore

func NewChore

func NewChore(log *zap.Logger, config Config, loop *segmentloop.Service) *Chore

NewChore creates a new instance of the metrics chore.

func (*Chore) Close

func (chore *Chore) Close() error

Close closes metrics chore.

func (*Chore) Run

func (chore *Chore) Run(ctx context.Context) (err error)

Run starts the metrics chore.

type Config

type Config struct {
	UseRangedLoop bool `help:"whether to use ranged loop instead of segment loop" default:"false"`
}

Config contains configurable values for metrics collection.

type Counter

type Counter struct {
	// number of objects that has at least one remote segment
	RemoteObjects int64
	// number of objects that has all inline segments
	InlineObjects int64

	// encrypted size
	TotalInlineBytes int64
	// encrypted size
	TotalRemoteBytes int64

	TotalInlineSegments int64
	TotalRemoteSegments int64
	// contains filtered or unexported fields
}

Counter implements the segment loop observer interface for data science metrics collection.

architecture: Observer

func NewCounter

func NewCounter() *Counter

NewCounter instantiates a new counter to be subscribed to the metainfo loop.

func (*Counter) InlineSegment

func (counter *Counter) InlineSegment(ctx context.Context, segment *segmentloop.Segment) error

InlineSegment increments the count for inline objects.

func (*Counter) LoopStarted

func (counter *Counter) LoopStarted(context.Context, segmentloop.LoopInfo) (err error)

LoopStarted is called at each start of a loop.

func (*Counter) RemoteSegment

func (counter *Counter) RemoteSegment(ctx context.Context, segment *segmentloop.Segment) error

RemoteSegment increments the count for objects with remote segments.

type Metrics

type Metrics struct {
	// RemoteObjects is the count of objects with at least one remote segment.
	RemoteObjects int64

	// InlineObjects is the count of objects with only inline segments.
	InlineObjects int64

	// TotalInlineBytes is the amount of bytes across all inline segments.
	TotalInlineBytes int64

	// TotalRemoteBytes is the amount of bytes across all remote segments.
	TotalRemoteBytes int64

	// TotalInlineSegments is the count of inline segments across all objects.
	TotalInlineSegments int64

	// TotalRemoteSegments is the count of remote segments across all objects.
	TotalRemoteSegments int64
}

Metrics represents the metrics that are tracked by this package.

func (*Metrics) Aggregate

func (metrics *Metrics) Aggregate(partial Metrics)

Aggregate aggregates the given metrics into the receiver.

func (*Metrics) Reset

func (metrics *Metrics) Reset()

Reset resets the invidual metrics back to zero.

type Observer

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

Observer implements the ranged segment loop observer interface for data science metrics collection.

func NewObserver

func NewObserver() *Observer

NewObserver instantiates a new rangedloop observer which aggregates object statistics from observed segments.

func (*Observer) Finish

func (obs *Observer) Finish(ctx context.Context) error

Finish emits the aggregated metrics.

func (*Observer) Fork

func (obs *Observer) Fork(ctx context.Context) (rangedloop.Partial, error)

Fork implements the Observer method of the same name by returning a partial implementation that aggregates metrics about the observed segments/streams. These metrics will be aggregated into the observed totals during Join.

func (*Observer) Join

func (obs *Observer) Join(ctx context.Context, partial rangedloop.Partial) error

Join aggregates the partial metrics.

func (*Observer) Start

func (obs *Observer) Start(ctx context.Context, startTime time.Time) error

Start implements the Observer method of the same name by resetting the aggregated metrics.

func (*Observer) TestingMetrics

func (obs *Observer) TestingMetrics() Metrics

TestingMetrics returns the accumulated metrics. It is intended to be called from tests.