TrueCloudLab/distribution
14
0
Fork 4
Code Issues 1 Pull requests 1 Projects Releases 2 Packages Wiki Activity Actions

vendor: github.com/prometheus/client_golang v1.12.1

Browse source 
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
This commit is contained in:
Sebastiaan van Stijn 2022-05-04 20:54:58 +02:00
parent 985711c1f4
commit ec47096efc
No known key found for this signature in database
GPG key ID: 76698F39D527CE8C
574 changed files with 101741 additions and 22828 deletions
Download patch file Download diff file Expand all files Collapse all files

174
vendor/github.com/prometheus/client_golang/prometheus/histogram.go generated vendored
View file

@ -20,7 +20,9 @@ import (
"sort"
"sync"
"sync/atomic"
"time"
//nolint:staticcheck // Ignore SA1019. Need to keep deprecated package for compatibility.
"github.com/golang/protobuf/proto"
dto "github.com/prometheus/client_model/go"
@ -45,7 +47,12 @@ type Histogram interface {
Metric
Collector
// Observe adds a single observation to the histogram.
// Observe adds a single observation to the histogram. Observations are
// usually positive or zero. Negative observations are accepted but
// prevent current versions of Prometheus from properly detecting
// counter resets in the sum of observations. See
// https://prometheus.io/docs/practices/histograms/#count-and-sum-of-observations
// for details.
Observe(float64)
}
@ -109,6 +116,34 @@ func ExponentialBuckets(start, factor float64, count int) []float64 {
return buckets
}
// ExponentialBucketsRange creates 'count' buckets, where the lowest bucket is
// 'min' and the highest bucket is 'max'. The final +Inf bucket is not counted
// and not included in the returned slice. The returned slice is meant to be
// used for the Buckets field of HistogramOpts.
//
// The function panics if 'count' is 0 or negative, if 'min' is 0 or negative.
func ExponentialBucketsRange(min, max float64, count int) []float64 {
if count < 1 {
panic("ExponentialBucketsRange count needs a positive count")
}
if min <= 0 {
panic("ExponentialBucketsRange min needs to be greater than 0")
}
// Formula for exponential buckets.
// max = min*growthFactor^(bucketCount-1)
// We know max/min and highest bucket. Solve for growthFactor.
growthFactor := math.Pow(max/min, 1.0/float64(count-1))
// Now that we know growthFactor, solve for each bucket.
buckets := make([]float64, count)
for i := 1; i <= count; i++ {
buckets[i-1] = min * math.Pow(growthFactor, float64(i-1))
}
return buckets
}
// HistogramOpts bundles the options for creating a Histogram metric. It is
// mandatory to set Name to a non-empty string. All other fields are optional
// and can safely be left at their zero value, although it is strongly
@ -138,7 +173,7 @@ type HistogramOpts struct {
// better covered by target labels set by the scraping Prometheus
// server, or by one specific metric (e.g. a build_info or a
// machine_role metric). See also
// https://prometheus.io/docs/instrumenting/writing_exporters/#target-labels,-not-static-scraped-labels
// https://prometheus.io/docs/instrumenting/writing_exporters/#target-labels-not-static-scraped-labels
ConstLabels Labels
// Buckets defines the buckets into which observations are counted. Each
@ -151,6 +186,10 @@ type HistogramOpts struct {
// NewHistogram creates a new Histogram based on the provided HistogramOpts. It
// panics if the buckets in HistogramOpts are not in strictly increasing order.
//
// The returned implementation also implements ExemplarObserver. It is safe to
// perform the corresponding type assertion. Exemplars are tracked separately
// for each bucket.
func NewHistogram(opts HistogramOpts) Histogram {
return newHistogram(
NewDesc(
@ -186,8 +225,9 @@ func newHistogram(desc *Desc, opts HistogramOpts, labelValues ...string) Histogr
h := &histogram{
desc: desc,
upperBounds: opts.Buckets,
labelPairs: makeLabelPairs(desc, labelValues),
counts: [2]*histogramCounts{&histogramCounts{}, &histogramCounts{}},
labelPairs: MakeLabelPairs(desc, labelValues),
counts: [2]*histogramCounts{{}, {}},
now: time.Now,
}
for i, upperBound := range h.upperBounds {
if i < len(h.upperBounds)-1 {
@ -205,9 +245,10 @@ func newHistogram(desc *Desc, opts HistogramOpts, labelValues ...string) Histogr
}
}
// Finally we know the final length of h.upperBounds and can make buckets
// for both counts:
// for both counts as well as exemplars:
h.counts[0].buckets = make([]uint64, len(h.upperBounds))
h.counts[1].buckets = make([]uint64, len(h.upperBounds))
h.exemplars = make([]atomic.Value, len(h.upperBounds)+1)
h.init(h) // Init self-collection.
return h
@ -254,6 +295,9 @@ type histogram struct {
upperBounds []float64
labelPairs []*dto.LabelPair
exemplars []atomic.Value // One more than buckets (to include +Inf), each a *dto.Exemplar.
now func() time.Time // To mock out time.Now() for testing.
}
func (h *histogram) Desc() *Desc {
@ -261,36 +305,13 @@ func (h *histogram) Desc() *Desc {
}
func (h *histogram) Observe(v float64) {
// TODO(beorn7): For small numbers of buckets (<30), a linear search is
// slightly faster than the binary search. If we really care, we could
// switch from one search strategy to the other depending on the number
// of buckets.
//
// Microbenchmarks (BenchmarkHistogramNoLabels):
// 11 buckets: 38.3 ns/op linear - binary 48.7 ns/op
// 100 buckets: 78.1 ns/op linear - binary 54.9 ns/op
// 300 buckets: 154 ns/op linear - binary 61.6 ns/op
i := sort.SearchFloat64s(h.upperBounds, v)
h.observe(v, h.findBucket(v))
}
// We increment h.countAndHotIdx so that the counter in the lower
// 63 bits gets incremented. At the same time, we get the new value
// back, which we can use to find the currently-hot counts.
n := atomic.AddUint64(&h.countAndHotIdx, 1)
hotCounts := h.counts[n>>63]
if i < len(h.upperBounds) {
atomic.AddUint64(&hotCounts.buckets[i], 1)
}
for {
oldBits := atomic.LoadUint64(&hotCounts.sumBits)
newBits := math.Float64bits(math.Float64frombits(oldBits) + v)
if atomic.CompareAndSwapUint64(&hotCounts.sumBits, oldBits, newBits) {
break
}
}
// Increment count last as we take it as a signal that the observation
// is complete.
atomic.AddUint64(&hotCounts.count, 1)
func (h *histogram) ObserveWithExemplar(v float64, e Labels) {
i := h.findBucket(v)
h.observe(v, i)
h.updateExemplar(v, i, e)
}
func (h *histogram) Write(out *dto.Metric) error {
@ -329,6 +350,18 @@ func (h *histogram) Write(out *dto.Metric) error {
CumulativeCount: proto.Uint64(cumCount),
UpperBound: proto.Float64(upperBound),
}
if e := h.exemplars[i].Load(); e != nil {
his.Bucket[i].Exemplar = e.(*dto.Exemplar)
}
}
// If there is an exemplar for the +Inf bucket, we have to add that bucket explicitly.
if e := h.exemplars[len(h.upperBounds)].Load(); e != nil {
b := &dto.Bucket{
CumulativeCount: proto.Uint64(count),
UpperBound: proto.Float64(math.Inf(1)),
Exemplar: e.(*dto.Exemplar),
}
his.Bucket = append(his.Bucket, b)
}
out.Histogram = his
@ -352,13 +385,64 @@ func (h *histogram) Write(out *dto.Metric) error {
return nil
}
// findBucket returns the index of the bucket for the provided value, or
// len(h.upperBounds) for the +Inf bucket.
func (h *histogram) findBucket(v float64) int {
// TODO(beorn7): For small numbers of buckets (<30), a linear search is
// slightly faster than the binary search. If we really care, we could
// switch from one search strategy to the other depending on the number
// of buckets.
//
// Microbenchmarks (BenchmarkHistogramNoLabels):
// 11 buckets: 38.3 ns/op linear - binary 48.7 ns/op
// 100 buckets: 78.1 ns/op linear - binary 54.9 ns/op
// 300 buckets: 154 ns/op linear - binary 61.6 ns/op
return sort.SearchFloat64s(h.upperBounds, v)
}
// observe is the implementation for Observe without the findBucket part.
func (h *histogram) observe(v float64, bucket int) {
// We increment h.countAndHotIdx so that the counter in the lower
// 63 bits gets incremented. At the same time, we get the new value
// back, which we can use to find the currently-hot counts.
n := atomic.AddUint64(&h.countAndHotIdx, 1)
hotCounts := h.counts[n>>63]
if bucket < len(h.upperBounds) {
atomic.AddUint64(&hotCounts.buckets[bucket], 1)
}
for {
oldBits := atomic.LoadUint64(&hotCounts.sumBits)
newBits := math.Float64bits(math.Float64frombits(oldBits) + v)
if atomic.CompareAndSwapUint64(&hotCounts.sumBits, oldBits, newBits) {
break
}
}
// Increment count last as we take it as a signal that the observation
// is complete.
atomic.AddUint64(&hotCounts.count, 1)
}
// updateExemplar replaces the exemplar for the provided bucket. With empty
// labels, it's a no-op. It panics if any of the labels is invalid.
func (h *histogram) updateExemplar(v float64, bucket int, l Labels) {
if l == nil {
return
}
e, err := newExemplar(v, h.now(), l)
if err != nil {
panic(err)
}
h.exemplars[bucket].Store(e)
}
// HistogramVec is a Collector that bundles a set of Histograms that all share the
// same Desc, but have different values for their variable labels. This is used
// if you want to count the same thing partitioned by various dimensions
// (e.g. HTTP request latencies, partitioned by status code and method). Create
// instances with NewHistogramVec.
type HistogramVec struct {
*metricVec
*MetricVec
}
// NewHistogramVec creates a new HistogramVec based on the provided HistogramOpts and
@ -371,14 +455,14 @@ func NewHistogramVec(opts HistogramOpts, labelNames []string) *HistogramVec {
opts.ConstLabels,
)
return &HistogramVec{
metricVec: newMetricVec(desc, func(lvs ...string) Metric {
MetricVec: NewMetricVec(desc, func(lvs ...string) Metric {
return newHistogram(desc, opts, lvs...)
}),
}
}
// GetMetricWithLabelValues returns the Histogram for the given slice of label
// values (same order as the VariableLabels in Desc). If that combination of
// values (same order as the variable labels in Desc). If that combination of
// label values is accessed for the first time, a new Histogram is created.
//
// It is possible to call this method without using the returned Histogram to only
@ -393,7 +477,7 @@ func NewHistogramVec(opts HistogramOpts, labelNames []string) *HistogramVec {
// example.
//
// An error is returned if the number of label values is not the same as the
// number of VariableLabels in Desc (minus any curried labels).
// number of variable labels in Desc (minus any curried labels).
//
// Note that for more than one label value, this method is prone to mistakes
// caused by an incorrect order of arguments. Consider GetMetricWith(Labels) as
@ -402,7 +486,7 @@ func NewHistogramVec(opts HistogramOpts, labelNames []string) *HistogramVec {
// with a performance overhead (for creating and processing the Labels map).
// See also the GaugeVec example.
func (v *HistogramVec) GetMetricWithLabelValues(lvs ...string) (Observer, error) {
metric, err := v.metricVec.getMetricWithLabelValues(lvs...)
metric, err := v.MetricVec.GetMetricWithLabelValues(lvs...)
if metric != nil {
return metric.(Observer), err
}
@ -410,19 +494,19 @@ func (v *HistogramVec) GetMetricWithLabelValues(lvs ...string) (Observer, error)
}
// GetMetricWith returns the Histogram for the given Labels map (the label names
// must match those of the VariableLabels in Desc). If that label map is
// must match those of the variable labels in Desc). If that label map is
// accessed for the first time, a new Histogram is created. Implications of
// creating a Histogram without using it and keeping the Histogram for later use
// are the same as for GetMetricWithLabelValues.
//
// An error is returned if the number and names of the Labels are inconsistent
// with those of the VariableLabels in Desc (minus any curried labels).
// with those of the variable labels in Desc (minus any curried labels).
//
// This method is used for the same purpose as
// GetMetricWithLabelValues(...string). See there for pros and cons of the two
// methods.
func (v *HistogramVec) GetMetricWith(labels Labels) (Observer, error) {
metric, err := v.metricVec.getMetricWith(labels)
metric, err := v.MetricVec.GetMetricWith(labels)
if metric != nil {
return metric.(Observer), err
}
@ -466,7 +550,7 @@ func (v *HistogramVec) With(labels Labels) Observer {
// registered with a given registry (usually the uncurried version). The Reset
// method deletes all metrics, even if called on a curried vector.
func (v *HistogramVec) CurryWith(labels Labels) (ObserverVec, error) {
vec, err := v.curryWith(labels)
vec, err := v.MetricVec.CurryWith(labels)
if vec != nil {
return &HistogramVec{vec}, err
}
@ -551,12 +635,12 @@ func NewConstHistogram(
count: count,
sum: sum,
buckets: buckets,
labelPairs: makeLabelPairs(desc, labelValues),
labelPairs: MakeLabelPairs(desc, labelValues),
}, nil
}
// MustNewConstHistogram is a version of NewConstHistogram that panics where
// NewConstMetric would have returned an error.
// NewConstHistogram would have returned an error.
func MustNewConstHistogram(
desc *Desc,
count uint64,