mirror of
https://github.com/prometheus/prometheus.git
synced 2024-12-28 15:09:39 -08:00
b3e34c6658
This commit introduces to Prometheus a batch database sample curator, which corroborates the high watermarks for sample series against the curation watermark table to see whether a curator of a given type needs to be run. The curator is an abstract executor, which runs various curation strategies across the database. It remarks the progress for each type of curation processor that runs for a given sample series. A curation procesor is responsible for effectuating the underlying batch changes that are request. In this commit, we introduce the CompactionProcessor, which takes several bits of runtime metadata and combine sparse sample entries in the database together to form larger groups. For instance, for a given series it would be possible to have the curator effectuate the following grouping: - Samples Older than Two Weeks: Grouped into Bunches of 10000 - Samples Older than One Week: Grouped into Bunches of 1000 - Samples Older than One Day: Grouped into Bunches of 100 - Samples Older than One Hour: Grouped into Bunches of 10 The benefits hereof of such a compaction are 1. a smaller search space in the database keyspace, 2. better employment of compression for repetious values, and 3. reduced seek times.
248 lines
5.4 KiB
Go
248 lines
5.4 KiB
Go
// Copyright 2013 Prometheus Team
|
|
// Licensed under the Apache License, Version 2.0 (the "License");
|
|
// you may not use this file except in compliance with the License.
|
|
// You may obtain a copy of the License at
|
|
//
|
|
// http://www.apache.org/licenses/LICENSE-2.0
|
|
//
|
|
// Unless required by applicable law or agreed to in writing, software
|
|
// distributed under the License is distributed on an "AS IS" BASIS,
|
|
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
// See the License for the specific language governing permissions and
|
|
// limitations under the License.
|
|
|
|
package model
|
|
|
|
import (
|
|
"bytes"
|
|
"code.google.com/p/goprotobuf/proto"
|
|
"fmt"
|
|
dto "github.com/prometheus/prometheus/model/generated"
|
|
"sort"
|
|
"time"
|
|
)
|
|
|
|
const (
|
|
// XXX: Re-evaluate down the road.
|
|
reservedDelimiter = `"`
|
|
)
|
|
|
|
// A LabelSet is a collection of LabelName and LabelValue pairs. The LabelSet
|
|
// may be fully-qualified down to the point where it may resolve to a single
|
|
// Metric in the data store or not. All operations that occur within the realm
|
|
// of a LabelSet can emit a vector of Metric entities to which the LabelSet may
|
|
// match.
|
|
type LabelSet map[LabelName]LabelValue
|
|
|
|
func (l LabelSet) String() string {
|
|
var (
|
|
buffer bytes.Buffer
|
|
labels LabelNames
|
|
labelCount int = len(l)
|
|
)
|
|
|
|
for name := range l {
|
|
labels = append(labels, name)
|
|
}
|
|
|
|
sort.Sort(labels)
|
|
|
|
fmt.Fprintf(&buffer, "{")
|
|
for i := 0; i < labelCount; i++ {
|
|
var (
|
|
label = labels[i]
|
|
value = l[label]
|
|
)
|
|
|
|
switch i {
|
|
case labelCount - 1:
|
|
fmt.Fprintf(&buffer, "%s=%s", label, value)
|
|
default:
|
|
fmt.Fprintf(&buffer, "%s=%s, ", label, value)
|
|
}
|
|
}
|
|
|
|
fmt.Fprintf(&buffer, "}")
|
|
|
|
return buffer.String()
|
|
}
|
|
|
|
func (l LabelSet) ToMetric() (metric Metric) {
|
|
metric = Metric{}
|
|
for label, value := range l {
|
|
metric[label] = value
|
|
}
|
|
return
|
|
}
|
|
|
|
// A Metric is similar to a LabelSet, but the key difference is that a Metric is
|
|
// a singleton and refers to one and only one stream of samples.
|
|
type Metric map[LabelName]LabelValue
|
|
|
|
// A SampleValue is a representation of a value for a given sample at a given
|
|
// time.
|
|
type SampleValue float64
|
|
|
|
func (s SampleValue) Equal(o SampleValue) bool {
|
|
return s == o
|
|
}
|
|
|
|
func (s SampleValue) ToDTO() *float64 {
|
|
return proto.Float64(float64(s))
|
|
}
|
|
|
|
func (v SampleValue) MarshalJSON() ([]byte, error) {
|
|
return []byte(fmt.Sprintf(`"%f"`, v)), nil
|
|
}
|
|
|
|
func (v SampleValue) String() string {
|
|
return fmt.Sprint(float64(v))
|
|
}
|
|
|
|
func (s SamplePair) MarshalJSON() ([]byte, error) {
|
|
return []byte(fmt.Sprintf("{\"Value\": \"%f\", \"Timestamp\": %d}", s.Value, s.Timestamp.Unix())), nil
|
|
}
|
|
|
|
type SamplePair struct {
|
|
Value SampleValue
|
|
Timestamp time.Time
|
|
}
|
|
|
|
func (s SamplePair) Equal(o SamplePair) bool {
|
|
return s.Value.Equal(o.Value) && s.Timestamp.Equal(o.Timestamp)
|
|
}
|
|
|
|
func (s SamplePair) ToDTO() (out *dto.SampleValueSeries_Value) {
|
|
out = &dto.SampleValueSeries_Value{
|
|
Timestamp: proto.Int64(s.Timestamp.Unix()),
|
|
Value: s.Value.ToDTO(),
|
|
}
|
|
|
|
return
|
|
}
|
|
|
|
func (s SamplePair) String() string {
|
|
return fmt.Sprintf("SamplePair at %s of %s", s.Timestamp, s.Value)
|
|
}
|
|
|
|
type Values []SamplePair
|
|
|
|
func (v Values) Len() int {
|
|
return len(v)
|
|
}
|
|
|
|
func (v Values) Less(i, j int) bool {
|
|
return v[i].Timestamp.Before(v[j].Timestamp)
|
|
}
|
|
|
|
func (v Values) Swap(i, j int) {
|
|
v[i], v[j] = v[j], v[i]
|
|
}
|
|
|
|
// FirstTimeAfter indicates whether the first sample of a set is after a given
|
|
// timestamp.
|
|
func (v Values) FirstTimeAfter(t time.Time) bool {
|
|
return v[0].Timestamp.After(t)
|
|
}
|
|
|
|
// LastTimeBefore indicates whether the last sample of a set is before a given
|
|
// timestamp.
|
|
func (v Values) LastTimeBefore(t time.Time) bool {
|
|
return v[len(v)-1].Timestamp.Before(t)
|
|
}
|
|
|
|
// InsideInterval indicates whether a given range of sorted values could contain
|
|
// a value for a given time.
|
|
func (v Values) InsideInterval(t time.Time) (s bool) {
|
|
if v.Len() == 0 {
|
|
return
|
|
}
|
|
|
|
if t.Before(v[0].Timestamp) {
|
|
return
|
|
}
|
|
|
|
if !v[v.Len()-1].Timestamp.Before(t) {
|
|
return
|
|
}
|
|
|
|
return true
|
|
}
|
|
|
|
// TruncateBefore returns a subslice of the original such that extraneous
|
|
// samples in the collection that occur before the provided time are
|
|
// dropped. The original slice is not mutated.
|
|
func (v Values) TruncateBefore(t time.Time) (values Values) {
|
|
index := sort.Search(len(v), func(i int) bool {
|
|
timestamp := v[i].Timestamp
|
|
|
|
return !timestamp.Before(t)
|
|
})
|
|
|
|
switch index {
|
|
case 0:
|
|
values = v
|
|
case len(v):
|
|
values = v[len(v)-1:]
|
|
default:
|
|
values = v[index-1:]
|
|
}
|
|
|
|
return
|
|
}
|
|
|
|
func (v Values) ToDTO() (out *dto.SampleValueSeries) {
|
|
out = &dto.SampleValueSeries{}
|
|
|
|
for _, value := range v {
|
|
out.Value = append(out.Value, value.ToDTO())
|
|
}
|
|
|
|
return
|
|
}
|
|
|
|
func (v Values) ToSampleKey(f Fingerprint) SampleKey {
|
|
return SampleKey{
|
|
Fingerprint: f,
|
|
FirstTimestamp: v[0].Timestamp,
|
|
LastTimestamp: v[len(v)-1].Timestamp,
|
|
SampleCount: uint32(len(v)),
|
|
}
|
|
}
|
|
|
|
func (v Values) String() string {
|
|
buffer := bytes.Buffer{}
|
|
|
|
fmt.Fprintf(&buffer, "[")
|
|
for i, value := range v {
|
|
fmt.Fprintf(&buffer, "%d. %s", i, value)
|
|
if i != len(v)-1 {
|
|
fmt.Fprintf(&buffer, "\n")
|
|
}
|
|
}
|
|
fmt.Fprintf(&buffer, "]")
|
|
|
|
return buffer.String()
|
|
}
|
|
|
|
func NewValuesFromDTO(dto *dto.SampleValueSeries) (v Values) {
|
|
for _, value := range dto.Value {
|
|
v = append(v, SamplePair{
|
|
Timestamp: time.Unix(*value.Timestamp, 0).UTC(),
|
|
Value: SampleValue(*value.Value),
|
|
})
|
|
}
|
|
|
|
return
|
|
}
|
|
|
|
type SampleSet struct {
|
|
Metric Metric
|
|
Values Values
|
|
}
|
|
|
|
type Interval struct {
|
|
OldestInclusive time.Time
|
|
NewestInclusive time.Time
|
|
}
|