prometheus/rules/ast/persistence_adapter.go

199 lines
6.2 KiB
Go
Raw Normal View History

// 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 ast
import (
"flag"
"time"
clientmodel "github.com/prometheus/client_golang/model"
"github.com/prometheus/prometheus/stats"
"github.com/prometheus/prometheus/storage/metric"
)
var defaultStalenessDelta = flag.Int("defaultStalenessDelta", 300, "Default staleness delta allowance in seconds during expression evaluations.")
// StalenessPolicy describes the lenience limits to apply to values
// from the materialized view.
type StalenessPolicy struct {
// Describes the inclusive limit at which individual points if requested will
// be matched and subject to interpolation.
DeltaAllowance time.Duration
}
type viewAdapter struct {
2013-05-07 11:15:10 +00:00
// Policy that dictates when sample values around an evaluation time are to
// be interpreted as stale.
stalenessPolicy StalenessPolicy
2013-05-07 11:15:10 +00:00
// AST-global storage to use for operations that are not supported by views
// (i.e. fingerprint->metric lookups).
storage *metric.TieredStorage
// The materialized view which contains all timeseries data required for
// executing a query.
view metric.View
// The TimerGroup object in which to capture query timing statistics.
stats *stats.TimerGroup
}
// interpolateSamples interpolates a value at a target time between two
// provided sample pairs.
Use custom timestamp type for sample timestamps and related code. So far we've been using Go's native time.Time for anything related to sample timestamps. Since the range of time.Time is much bigger than what we need, this has created two problems: - there could be time.Time values which were out of the range/precision of the time type that we persist to disk, therefore causing incorrectly ordered keys. One bug caused by this was: https://github.com/prometheus/prometheus/issues/367 It would be good to use a timestamp type that's more closely aligned with what the underlying storage supports. - sizeof(time.Time) is 192, while Prometheus should be ok with a single 64-bit Unix timestamp (possibly even a 32-bit one). Since we store samples in large numbers, this seriously affects memory usage. Furthermore, copying/working with the data will be faster if it's smaller. *MEMORY USAGE RESULTS* Initial memory usage comparisons for a running Prometheus with 1 timeseries and 100,000 samples show roughly a 13% decrease in total (VIRT) memory usage. In my tests, this advantage for some reason decreased a bit the more samples the timeseries had (to 5-7% for millions of samples). This I can't fully explain, but perhaps garbage collection issues were involved. *WHEN TO USE THE NEW TIMESTAMP TYPE* The new clientmodel.Timestamp type should be used whenever time calculations are either directly or indirectly related to sample timestamps. For example: - the timestamp of a sample itself - all kinds of watermarks - anything that may become or is compared to a sample timestamp (like the timestamp passed into Target.Scrape()). When to still use time.Time: - for measuring durations/times not related to sample timestamps, like duration telemetry exporting, timers that indicate how frequently to execute some action, etc. *NOTE ON OPERATOR OPTIMIZATION TESTS* We don't use operator optimization code anymore, but it still lives in the code as dead code. It still has tests, but I couldn't get all of them to pass with the new timestamp format. I commented out the failing cases for now, but we should probably remove the dead code soon. I just didn't want to do that in the same change as this. Change-Id: I821787414b0debe85c9fffaeb57abd453727af0f
2013-10-28 13:35:02 +00:00
func interpolateSamples(first, second *metric.SamplePair, timestamp clientmodel.Timestamp) *metric.SamplePair {
dv := second.Value - first.Value
dt := second.Timestamp.Sub(first.Timestamp)
dDt := dv / clientmodel.SampleValue(dt)
offset := clientmodel.SampleValue(timestamp.Sub(first.Timestamp))
return &metric.SamplePair{
Value: first.Value + (offset * dDt),
Timestamp: timestamp,
}
}
// chooseClosestSample chooses the closest sample of a list of samples
// surrounding a given target time. If samples are found both before and after
// the target time, the sample value is interpolated between these. Otherwise,
// the single closest sample is returned verbatim.
Use custom timestamp type for sample timestamps and related code. So far we've been using Go's native time.Time for anything related to sample timestamps. Since the range of time.Time is much bigger than what we need, this has created two problems: - there could be time.Time values which were out of the range/precision of the time type that we persist to disk, therefore causing incorrectly ordered keys. One bug caused by this was: https://github.com/prometheus/prometheus/issues/367 It would be good to use a timestamp type that's more closely aligned with what the underlying storage supports. - sizeof(time.Time) is 192, while Prometheus should be ok with a single 64-bit Unix timestamp (possibly even a 32-bit one). Since we store samples in large numbers, this seriously affects memory usage. Furthermore, copying/working with the data will be faster if it's smaller. *MEMORY USAGE RESULTS* Initial memory usage comparisons for a running Prometheus with 1 timeseries and 100,000 samples show roughly a 13% decrease in total (VIRT) memory usage. In my tests, this advantage for some reason decreased a bit the more samples the timeseries had (to 5-7% for millions of samples). This I can't fully explain, but perhaps garbage collection issues were involved. *WHEN TO USE THE NEW TIMESTAMP TYPE* The new clientmodel.Timestamp type should be used whenever time calculations are either directly or indirectly related to sample timestamps. For example: - the timestamp of a sample itself - all kinds of watermarks - anything that may become or is compared to a sample timestamp (like the timestamp passed into Target.Scrape()). When to still use time.Time: - for measuring durations/times not related to sample timestamps, like duration telemetry exporting, timers that indicate how frequently to execute some action, etc. *NOTE ON OPERATOR OPTIMIZATION TESTS* We don't use operator optimization code anymore, but it still lives in the code as dead code. It still has tests, but I couldn't get all of them to pass with the new timestamp format. I commented out the failing cases for now, but we should probably remove the dead code soon. I just didn't want to do that in the same change as this. Change-Id: I821787414b0debe85c9fffaeb57abd453727af0f
2013-10-28 13:35:02 +00:00
func (v *viewAdapter) chooseClosestSample(samples metric.Values, timestamp clientmodel.Timestamp) *metric.SamplePair {
var closestBefore *metric.SamplePair
var closestAfter *metric.SamplePair
for _, candidate := range samples {
delta := candidate.Timestamp.Sub(timestamp)
// Samples before target time.
if delta < 0 {
// Ignore samples outside of staleness policy window.
if -delta > v.stalenessPolicy.DeltaAllowance {
continue
}
// Ignore samples that are farther away than what we've seen before.
if closestBefore != nil && candidate.Timestamp.Before(closestBefore.Timestamp) {
continue
}
sample := candidate
closestBefore = sample
}
// Samples after target time.
if delta >= 0 {
// Ignore samples outside of staleness policy window.
if delta > v.stalenessPolicy.DeltaAllowance {
continue
}
// Ignore samples that are farther away than samples we've seen before.
if closestAfter != nil && candidate.Timestamp.After(closestAfter.Timestamp) {
continue
}
sample := candidate
closestAfter = sample
}
}
switch {
case closestBefore != nil && closestAfter != nil:
return interpolateSamples(closestBefore, closestAfter, timestamp)
case closestBefore != nil:
return closestBefore
default:
return closestAfter
}
}
Use custom timestamp type for sample timestamps and related code. So far we've been using Go's native time.Time for anything related to sample timestamps. Since the range of time.Time is much bigger than what we need, this has created two problems: - there could be time.Time values which were out of the range/precision of the time type that we persist to disk, therefore causing incorrectly ordered keys. One bug caused by this was: https://github.com/prometheus/prometheus/issues/367 It would be good to use a timestamp type that's more closely aligned with what the underlying storage supports. - sizeof(time.Time) is 192, while Prometheus should be ok with a single 64-bit Unix timestamp (possibly even a 32-bit one). Since we store samples in large numbers, this seriously affects memory usage. Furthermore, copying/working with the data will be faster if it's smaller. *MEMORY USAGE RESULTS* Initial memory usage comparisons for a running Prometheus with 1 timeseries and 100,000 samples show roughly a 13% decrease in total (VIRT) memory usage. In my tests, this advantage for some reason decreased a bit the more samples the timeseries had (to 5-7% for millions of samples). This I can't fully explain, but perhaps garbage collection issues were involved. *WHEN TO USE THE NEW TIMESTAMP TYPE* The new clientmodel.Timestamp type should be used whenever time calculations are either directly or indirectly related to sample timestamps. For example: - the timestamp of a sample itself - all kinds of watermarks - anything that may become or is compared to a sample timestamp (like the timestamp passed into Target.Scrape()). When to still use time.Time: - for measuring durations/times not related to sample timestamps, like duration telemetry exporting, timers that indicate how frequently to execute some action, etc. *NOTE ON OPERATOR OPTIMIZATION TESTS* We don't use operator optimization code anymore, but it still lives in the code as dead code. It still has tests, but I couldn't get all of them to pass with the new timestamp format. I commented out the failing cases for now, but we should probably remove the dead code soon. I just didn't want to do that in the same change as this. Change-Id: I821787414b0debe85c9fffaeb57abd453727af0f
2013-10-28 13:35:02 +00:00
func (v *viewAdapter) GetValueAtTime(fingerprints clientmodel.Fingerprints, timestamp clientmodel.Timestamp) (Vector, error) {
timer := v.stats.GetTimer(stats.GetValueAtTimeTime).Start()
samples := Vector{}
for _, fingerprint := range fingerprints {
sampleCandidates := v.view.GetValueAtTime(fingerprint, timestamp)
samplePair := v.chooseClosestSample(sampleCandidates, timestamp)
2013-05-07 11:15:10 +00:00
m, err := v.storage.GetMetricForFingerprint(fingerprint)
if err != nil {
return nil, err
}
if samplePair != nil {
samples = append(samples, &clientmodel.Sample{
Metric: m,
Value: samplePair.Value,
Timestamp: timestamp,
})
}
}
timer.Stop()
return samples, nil
}
func (v *viewAdapter) GetBoundaryValues(fingerprints clientmodel.Fingerprints, interval *metric.Interval) ([]metric.SampleSet, error) {
timer := v.stats.GetTimer(stats.GetBoundaryValuesTime).Start()
sampleSets := []metric.SampleSet{}
for _, fingerprint := range fingerprints {
samplePairs := v.view.GetBoundaryValues(fingerprint, *interval)
if len(samplePairs) == 0 {
continue
}
// TODO: memoize/cache this.
2013-05-07 11:15:10 +00:00
m, err := v.storage.GetMetricForFingerprint(fingerprint)
if err != nil {
return nil, err
}
sampleSet := metric.SampleSet{
Metric: m,
Values: samplePairs,
}
sampleSets = append(sampleSets, sampleSet)
}
timer.Stop()
return sampleSets, nil
}
func (v *viewAdapter) GetRangeValues(fingerprints clientmodel.Fingerprints, interval *metric.Interval) ([]metric.SampleSet, error) {
timer := v.stats.GetTimer(stats.GetRangeValuesTime).Start()
sampleSets := []metric.SampleSet{}
for _, fingerprint := range fingerprints {
samplePairs := v.view.GetRangeValues(fingerprint, *interval)
if len(samplePairs) == 0 {
continue
}
// TODO: memoize/cache this.
2013-05-07 11:15:10 +00:00
m, err := v.storage.GetMetricForFingerprint(fingerprint)
if err != nil {
return nil, err
}
sampleSet := metric.SampleSet{
Metric: m,
Values: samplePairs,
}
sampleSets = append(sampleSets, sampleSet)
}
timer.Stop()
return sampleSets, nil
}
// NewViewAdapter returns an initialized view adapter with a default
// staleness policy (based on the --defaultStalenessDelta flag).
func NewViewAdapter(view metric.View, storage *metric.TieredStorage, queryStats *stats.TimerGroup) *viewAdapter {
stalenessPolicy := StalenessPolicy{
DeltaAllowance: time.Duration(*defaultStalenessDelta) * time.Second,
}
return &viewAdapter{
stalenessPolicy: stalenessPolicy,
2013-05-07 11:15:10 +00:00
storage: storage,
view: view,
stats: queryStats,
}
}