prometheus/web/federate_test.go

420 lines
12 KiB
Go
Raw Normal View History

// Copyright 2016 The Prometheus Authors
// 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 web
import (
"bytes"
"context"
"fmt"
"io"
"net/http"
"net/http/httptest"
"sort"
"strings"
"testing"
"time"
"github.com/pkg/errors"
"github.com/prometheus/common/model"
"github.com/stretchr/testify/require"
"github.com/prometheus/prometheus/config"
"github.com/prometheus/prometheus/model/histogram"
"github.com/prometheus/prometheus/model/labels"
"github.com/prometheus/prometheus/model/textparse"
"github.com/prometheus/prometheus/promql"
"github.com/prometheus/prometheus/storage"
"github.com/prometheus/prometheus/tsdb"
)
var scenarios = map[string]struct {
params string
externalLabels labels.Labels
code int
body string
}{
"empty": {
params: "",
code: 200,
body: ``,
},
"match nothing": {
params: "match[]=does_not_match_anything",
code: 200,
body: ``,
},
"invalid params from the beginning": {
params: "match[]=-not-a-valid-metric-name",
code: 400,
body: `1:1: parse error: unexpected <op:->
`,
},
"invalid params somewhere in the middle": {
params: "match[]=not-a-valid-metric-name",
code: 400,
body: `1:4: parse error: unexpected <op:->
`,
},
"test_metric1": {
params: "match[]=test_metric1",
code: 200,
body: `# TYPE test_metric1 untyped
test_metric1{foo="bar",instance="i"} 10000 6000000
test_metric1{foo="boo",instance="i"} 1 6000000
`,
},
"test_metric2": {
params: "match[]=test_metric2",
code: 200,
body: `# TYPE test_metric2 untyped
test_metric2{foo="boo",instance="i"} 1 6000000
`,
},
"test_metric_without_labels": {
params: "match[]=test_metric_without_labels",
code: 200,
body: `# TYPE test_metric_without_labels untyped
test_metric_without_labels{instance=""} 1001 6000000
`,
},
"test_stale_metric": {
params: "match[]=test_metric_stale",
code: 200,
body: ``,
},
"test_old_metric": {
params: "match[]=test_metric_old",
code: 200,
body: `# TYPE test_metric_old untyped
test_metric_old{instance=""} 981 5880000
`,
},
"{foo='boo'}": {
params: "match[]={foo='boo'}",
code: 200,
body: `# TYPE test_metric1 untyped
test_metric1{foo="boo",instance="i"} 1 6000000
# TYPE test_metric2 untyped
test_metric2{foo="boo",instance="i"} 1 6000000
`,
},
"two matchers": {
params: "match[]=test_metric1&match[]=test_metric2",
code: 200,
body: `# TYPE test_metric1 untyped
test_metric1{foo="bar",instance="i"} 10000 6000000
test_metric1{foo="boo",instance="i"} 1 6000000
# TYPE test_metric2 untyped
test_metric2{foo="boo",instance="i"} 1 6000000
`,
},
"two matchers with overlap": {
params: "match[]={__name__=~'test_metric1'}&match[]={foo='bar'}",
code: 200,
body: `# TYPE test_metric1 untyped
test_metric1{foo="bar",instance="i"} 10000 6000000
test_metric1{foo="boo",instance="i"} 1 6000000
`,
},
"everything": {
params: "match[]={__name__=~'.%2b'}", // '%2b' is an URL-encoded '+'.
code: 200,
body: `# TYPE test_metric1 untyped
test_metric1{foo="bar",instance="i"} 10000 6000000
test_metric1{foo="boo",instance="i"} 1 6000000
# TYPE test_metric2 untyped
test_metric2{foo="boo",instance="i"} 1 6000000
# TYPE test_metric_old untyped
test_metric_old{instance=""} 981 5880000
# TYPE test_metric_without_labels untyped
test_metric_without_labels{instance=""} 1001 6000000
`,
},
"empty label value matches everything that doesn't have that label": {
params: "match[]={foo='',__name__=~'.%2b'}",
code: 200,
body: `# TYPE test_metric_old untyped
test_metric_old{instance=""} 981 5880000
# TYPE test_metric_without_labels untyped
test_metric_without_labels{instance=""} 1001 6000000
`,
},
"empty label value for a label that doesn't exist at all, matches everything": {
params: "match[]={bar='',__name__=~'.%2b'}",
code: 200,
body: `# TYPE test_metric1 untyped
test_metric1{foo="bar",instance="i"} 10000 6000000
test_metric1{foo="boo",instance="i"} 1 6000000
# TYPE test_metric2 untyped
test_metric2{foo="boo",instance="i"} 1 6000000
# TYPE test_metric_old untyped
test_metric_old{instance=""} 981 5880000
# TYPE test_metric_without_labels untyped
test_metric_without_labels{instance=""} 1001 6000000
`,
},
"external labels are added if not already present": {
params: "match[]={__name__=~'.%2b'}", // '%2b' is an URL-encoded '+'.
externalLabels: labels.FromStrings("foo", "baz", "zone", "ie"),
code: 200,
body: `# TYPE test_metric1 untyped
test_metric1{foo="bar",instance="i",zone="ie"} 10000 6000000
test_metric1{foo="boo",instance="i",zone="ie"} 1 6000000
# TYPE test_metric2 untyped
test_metric2{foo="boo",instance="i",zone="ie"} 1 6000000
# TYPE test_metric_old untyped
test_metric_old{foo="baz",instance="",zone="ie"} 981 5880000
# TYPE test_metric_without_labels untyped
test_metric_without_labels{foo="baz",instance="",zone="ie"} 1001 6000000
`,
},
"instance is an external label": {
// This makes no sense as a configuration, but we should
// know what it does anyway.
params: "match[]={__name__=~'.%2b'}", // '%2b' is an URL-encoded '+'.
externalLabels: labels.FromStrings("instance", "baz"),
code: 200,
body: `# TYPE test_metric1 untyped
test_metric1{foo="bar",instance="i"} 10000 6000000
test_metric1{foo="boo",instance="i"} 1 6000000
# TYPE test_metric2 untyped
test_metric2{foo="boo",instance="i"} 1 6000000
# TYPE test_metric_old untyped
test_metric_old{instance="baz"} 981 5880000
# TYPE test_metric_without_labels untyped
test_metric_without_labels{instance="baz"} 1001 6000000
`,
},
}
func TestFederation(t *testing.T) {
suite, err := promql.NewTest(t, `
load 1m
test_metric1{foo="bar",instance="i"} 0+100x100
test_metric1{foo="boo",instance="i"} 1+0x100
test_metric2{foo="boo",instance="i"} 1+0x100
test_metric_without_labels 1+10x100
test_metric_stale 1+10x99 stale
test_metric_old 1+10x98
`)
if err != nil {
t.Fatal(err)
}
defer suite.Close()
if err := suite.Run(); err != nil {
t.Fatal(err)
}
h := &Handler{
localStorage: &dbAdapter{suite.TSDB()},
lookbackDelta: 5 * time.Minute,
now: func() model.Time { return 101 * 60 * 1000 }, // 101min after epoch.
config: &config.Config{
GlobalConfig: config.GlobalConfig{},
},
}
for name, scenario := range scenarios {
t.Run(name, func(t *testing.T) {
h.config.GlobalConfig.ExternalLabels = scenario.externalLabels
req := httptest.NewRequest("GET", "http://example.org/federate?"+scenario.params, nil)
res := httptest.NewRecorder()
h.federation(res, req)
require.Equal(t, scenario.code, res.Code)
require.Equal(t, scenario.body, normalizeBody(res.Body))
})
}
}
type notReadyReadStorage struct {
LocalStorage
}
func (notReadyReadStorage) Querier(context.Context, int64, int64) (storage.Querier, error) {
return nil, errors.Wrap(tsdb.ErrNotReady, "wrap")
}
func (notReadyReadStorage) StartTime() (int64, error) {
return 0, errors.Wrap(tsdb.ErrNotReady, "wrap")
}
func (notReadyReadStorage) Stats(string) (*tsdb.Stats, error) {
return nil, errors.Wrap(tsdb.ErrNotReady, "wrap")
}
// Regression test for https://github.com/prometheus/prometheus/issues/7181.
func TestFederation_NotReady(t *testing.T) {
for name, scenario := range scenarios {
t.Run(name, func(t *testing.T) {
h := &Handler{
localStorage: notReadyReadStorage{},
lookbackDelta: 5 * time.Minute,
now: func() model.Time { return 101 * 60 * 1000 }, // 101min after epoch.
config: &config.Config{
GlobalConfig: config.GlobalConfig{
ExternalLabels: scenario.externalLabels,
},
},
}
req := httptest.NewRequest("GET", "http://example.org/federate?"+scenario.params, nil)
res := httptest.NewRecorder()
h.federation(res, req)
if scenario.code == http.StatusBadRequest {
// Request are expected to be checked before DB readiness.
require.Equal(t, http.StatusBadRequest, res.Code)
return
}
require.Equal(t, http.StatusServiceUnavailable, res.Code)
})
}
}
// normalizeBody sorts the lines within a metric to make it easy to verify the body.
// (Federation is not taking care of sorting within a metric family.)
func normalizeBody(body *bytes.Buffer) string {
var (
lines []string
lastHash int
)
for line, err := body.ReadString('\n'); err == nil; line, err = body.ReadString('\n') {
if line[0] == '#' && len(lines) > 0 {
sort.Strings(lines[lastHash+1:])
lastHash = len(lines)
}
lines = append(lines, line)
}
if len(lines) > 0 {
sort.Strings(lines[lastHash+1:])
}
return strings.Join(lines, "")
}
func TestFederationWithNativeHistograms(t *testing.T) {
suite, err := promql.NewTest(t, "")
if err != nil {
t.Fatal(err)
}
defer suite.Close()
if err := suite.Run(); err != nil {
t.Fatal(err)
}
var expVec promql.Vector
db := suite.TSDB()
hist := &histogram.Histogram{
Count: 10,
ZeroCount: 2,
ZeroThreshold: 0.001,
Sum: 39.4,
Schema: 1,
PositiveSpans: []histogram.Span{
{Offset: 0, Length: 2},
{Offset: 1, Length: 2},
},
PositiveBuckets: []int64{1, 1, -1, 0},
NegativeSpans: []histogram.Span{
{Offset: 0, Length: 2},
{Offset: 1, Length: 2},
},
NegativeBuckets: []int64{1, 1, -1, 0},
}
app := db.Appender(context.Background())
for i := 0; i < 6; i++ {
l := labels.FromStrings("__name__", "test_metric", "foo", fmt.Sprintf("%d", i))
expL := labels.FromStrings("__name__", "test_metric", "instance", "", "foo", fmt.Sprintf("%d", i))
if i%3 == 0 {
_, err = app.Append(0, l, 100*60*1000, float64(i*100))
expVec = append(expVec, promql.Sample{
promql: Separate `Point` into `FPoint` and `HPoint` In other words: Instead of having a “polymorphous” `Point` that can either contain a float value or a histogram value, use an `FPoint` for floats and an `HPoint` for histograms. This seemingly small change has a _lot_ of repercussions throughout the codebase. The idea here is to avoid the increase in size of `Point` arrays that happened after native histograms had been added. The higher-level data structures (`Sample`, `Series`, etc.) are still “polymorphous”. The same idea could be applied to them, but at each step the trade-offs needed to be evaluated. The idea with this change is to do the minimum necessary to get back to pre-histogram performance for functions that do not touch histograms. Here are comparisons for the `changes` function. The test data doesn't include histograms yet. Ideally, there would be no change in the benchmark result at all. First runtime v2.39 compared to directly prior to this commit: ``` name old time/op new time/op delta RangeQuery/expr=changes(a_one[1d]),steps=1-16 391µs ± 2% 542µs ± 1% +38.58% (p=0.000 n=9+8) RangeQuery/expr=changes(a_one[1d]),steps=10-16 452µs ± 2% 617µs ± 2% +36.48% (p=0.000 n=10+10) RangeQuery/expr=changes(a_one[1d]),steps=100-16 1.12ms ± 1% 1.36ms ± 2% +21.58% (p=0.000 n=8+10) RangeQuery/expr=changes(a_one[1d]),steps=1000-16 7.83ms ± 1% 8.94ms ± 1% +14.21% (p=0.000 n=10+10) RangeQuery/expr=changes(a_ten[1d]),steps=1-16 2.98ms ± 0% 3.30ms ± 1% +10.67% (p=0.000 n=9+10) RangeQuery/expr=changes(a_ten[1d]),steps=10-16 3.66ms ± 1% 4.10ms ± 1% +11.82% (p=0.000 n=10+10) RangeQuery/expr=changes(a_ten[1d]),steps=100-16 10.5ms ± 0% 11.8ms ± 1% +12.50% (p=0.000 n=8+10) RangeQuery/expr=changes(a_ten[1d]),steps=1000-16 77.6ms ± 1% 87.4ms ± 1% +12.63% (p=0.000 n=9+9) RangeQuery/expr=changes(a_hundred[1d]),steps=1-16 30.4ms ± 2% 32.8ms ± 1% +8.01% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=10-16 37.1ms ± 2% 40.6ms ± 2% +9.64% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=100-16 105ms ± 1% 117ms ± 1% +11.69% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=1000-16 783ms ± 3% 876ms ± 1% +11.83% (p=0.000 n=9+10) ``` And then runtime v2.39 compared to after this commit: ``` name old time/op new time/op delta RangeQuery/expr=changes(a_one[1d]),steps=1-16 391µs ± 2% 547µs ± 1% +39.84% (p=0.000 n=9+8) RangeQuery/expr=changes(a_one[1d]),steps=10-16 452µs ± 2% 616µs ± 2% +36.15% (p=0.000 n=10+10) RangeQuery/expr=changes(a_one[1d]),steps=100-16 1.12ms ± 1% 1.26ms ± 1% +12.20% (p=0.000 n=8+10) RangeQuery/expr=changes(a_one[1d]),steps=1000-16 7.83ms ± 1% 7.95ms ± 1% +1.59% (p=0.000 n=10+8) RangeQuery/expr=changes(a_ten[1d]),steps=1-16 2.98ms ± 0% 3.38ms ± 2% +13.49% (p=0.000 n=9+10) RangeQuery/expr=changes(a_ten[1d]),steps=10-16 3.66ms ± 1% 4.02ms ± 1% +9.80% (p=0.000 n=10+9) RangeQuery/expr=changes(a_ten[1d]),steps=100-16 10.5ms ± 0% 10.8ms ± 1% +3.08% (p=0.000 n=8+10) RangeQuery/expr=changes(a_ten[1d]),steps=1000-16 77.6ms ± 1% 78.1ms ± 1% +0.58% (p=0.035 n=9+10) RangeQuery/expr=changes(a_hundred[1d]),steps=1-16 30.4ms ± 2% 33.5ms ± 4% +10.18% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=10-16 37.1ms ± 2% 40.0ms ± 1% +7.98% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=100-16 105ms ± 1% 107ms ± 1% +1.92% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=1000-16 783ms ± 3% 775ms ± 1% -1.02% (p=0.019 n=9+9) ``` In summary, the runtime doesn't really improve with this change for queries with just a few steps. For queries with many steps, this commit essentially reinstates the old performance. This is good because the many-step queries are the one that matter most (longest absolute runtime). In terms of allocations, though, this commit doesn't make a dent at all (numbers not shown). The reason is that most of the allocations happen in the sampleRingIterator (in the storage package), which has to be addressed in a separate commit. Signed-off-by: beorn7 <beorn@grafana.com>
2022-10-28 07:58:40 -07:00
T: 100 * 60 * 1000,
F: float64(i * 100),
Metric: expL,
})
} else {
hist.ZeroCount++
_, err = app.AppendHistogram(0, l, 100*60*1000, hist.Copy(), nil)
expVec = append(expVec, promql.Sample{
promql: Separate `Point` into `FPoint` and `HPoint` In other words: Instead of having a “polymorphous” `Point` that can either contain a float value or a histogram value, use an `FPoint` for floats and an `HPoint` for histograms. This seemingly small change has a _lot_ of repercussions throughout the codebase. The idea here is to avoid the increase in size of `Point` arrays that happened after native histograms had been added. The higher-level data structures (`Sample`, `Series`, etc.) are still “polymorphous”. The same idea could be applied to them, but at each step the trade-offs needed to be evaluated. The idea with this change is to do the minimum necessary to get back to pre-histogram performance for functions that do not touch histograms. Here are comparisons for the `changes` function. The test data doesn't include histograms yet. Ideally, there would be no change in the benchmark result at all. First runtime v2.39 compared to directly prior to this commit: ``` name old time/op new time/op delta RangeQuery/expr=changes(a_one[1d]),steps=1-16 391µs ± 2% 542µs ± 1% +38.58% (p=0.000 n=9+8) RangeQuery/expr=changes(a_one[1d]),steps=10-16 452µs ± 2% 617µs ± 2% +36.48% (p=0.000 n=10+10) RangeQuery/expr=changes(a_one[1d]),steps=100-16 1.12ms ± 1% 1.36ms ± 2% +21.58% (p=0.000 n=8+10) RangeQuery/expr=changes(a_one[1d]),steps=1000-16 7.83ms ± 1% 8.94ms ± 1% +14.21% (p=0.000 n=10+10) RangeQuery/expr=changes(a_ten[1d]),steps=1-16 2.98ms ± 0% 3.30ms ± 1% +10.67% (p=0.000 n=9+10) RangeQuery/expr=changes(a_ten[1d]),steps=10-16 3.66ms ± 1% 4.10ms ± 1% +11.82% (p=0.000 n=10+10) RangeQuery/expr=changes(a_ten[1d]),steps=100-16 10.5ms ± 0% 11.8ms ± 1% +12.50% (p=0.000 n=8+10) RangeQuery/expr=changes(a_ten[1d]),steps=1000-16 77.6ms ± 1% 87.4ms ± 1% +12.63% (p=0.000 n=9+9) RangeQuery/expr=changes(a_hundred[1d]),steps=1-16 30.4ms ± 2% 32.8ms ± 1% +8.01% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=10-16 37.1ms ± 2% 40.6ms ± 2% +9.64% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=100-16 105ms ± 1% 117ms ± 1% +11.69% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=1000-16 783ms ± 3% 876ms ± 1% +11.83% (p=0.000 n=9+10) ``` And then runtime v2.39 compared to after this commit: ``` name old time/op new time/op delta RangeQuery/expr=changes(a_one[1d]),steps=1-16 391µs ± 2% 547µs ± 1% +39.84% (p=0.000 n=9+8) RangeQuery/expr=changes(a_one[1d]),steps=10-16 452µs ± 2% 616µs ± 2% +36.15% (p=0.000 n=10+10) RangeQuery/expr=changes(a_one[1d]),steps=100-16 1.12ms ± 1% 1.26ms ± 1% +12.20% (p=0.000 n=8+10) RangeQuery/expr=changes(a_one[1d]),steps=1000-16 7.83ms ± 1% 7.95ms ± 1% +1.59% (p=0.000 n=10+8) RangeQuery/expr=changes(a_ten[1d]),steps=1-16 2.98ms ± 0% 3.38ms ± 2% +13.49% (p=0.000 n=9+10) RangeQuery/expr=changes(a_ten[1d]),steps=10-16 3.66ms ± 1% 4.02ms ± 1% +9.80% (p=0.000 n=10+9) RangeQuery/expr=changes(a_ten[1d]),steps=100-16 10.5ms ± 0% 10.8ms ± 1% +3.08% (p=0.000 n=8+10) RangeQuery/expr=changes(a_ten[1d]),steps=1000-16 77.6ms ± 1% 78.1ms ± 1% +0.58% (p=0.035 n=9+10) RangeQuery/expr=changes(a_hundred[1d]),steps=1-16 30.4ms ± 2% 33.5ms ± 4% +10.18% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=10-16 37.1ms ± 2% 40.0ms ± 1% +7.98% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=100-16 105ms ± 1% 107ms ± 1% +1.92% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=1000-16 783ms ± 3% 775ms ± 1% -1.02% (p=0.019 n=9+9) ``` In summary, the runtime doesn't really improve with this change for queries with just a few steps. For queries with many steps, this commit essentially reinstates the old performance. This is good because the many-step queries are the one that matter most (longest absolute runtime). In terms of allocations, though, this commit doesn't make a dent at all (numbers not shown). The reason is that most of the allocations happen in the sampleRingIterator (in the storage package), which has to be addressed in a separate commit. Signed-off-by: beorn7 <beorn@grafana.com>
2022-10-28 07:58:40 -07:00
T: 100 * 60 * 1000,
H: hist.ToFloat(),
Metric: expL,
})
}
require.NoError(t, err)
}
require.NoError(t, app.Commit())
h := &Handler{
localStorage: &dbAdapter{suite.TSDB()},
lookbackDelta: 5 * time.Minute,
now: func() model.Time { return 101 * 60 * 1000 }, // 101min after epoch.
config: &config.Config{
GlobalConfig: config.GlobalConfig{},
},
}
req := httptest.NewRequest("GET", "http://example.org/federate?match[]=test_metric", nil)
req.Header.Add("Accept", `application/vnd.google.protobuf;proto=io.prometheus.client.MetricFamily;encoding=delimited,application/openmetrics-text;version=1.0.0;q=0.8,application/openmetrics-text;version=0.0.1;q=0.75,text/plain;version=0.0.4;q=0.5,*/*;q=0.1`)
res := httptest.NewRecorder()
h.federation(res, req)
require.Equal(t, http.StatusOK, res.Code)
body, err := io.ReadAll(res.Body)
require.NoError(t, err)
p := textparse.NewProtobufParser(body)
var actVec promql.Vector
metricFamilies := 0
promql: Separate `Point` into `FPoint` and `HPoint` In other words: Instead of having a “polymorphous” `Point` that can either contain a float value or a histogram value, use an `FPoint` for floats and an `HPoint` for histograms. This seemingly small change has a _lot_ of repercussions throughout the codebase. The idea here is to avoid the increase in size of `Point` arrays that happened after native histograms had been added. The higher-level data structures (`Sample`, `Series`, etc.) are still “polymorphous”. The same idea could be applied to them, but at each step the trade-offs needed to be evaluated. The idea with this change is to do the minimum necessary to get back to pre-histogram performance for functions that do not touch histograms. Here are comparisons for the `changes` function. The test data doesn't include histograms yet. Ideally, there would be no change in the benchmark result at all. First runtime v2.39 compared to directly prior to this commit: ``` name old time/op new time/op delta RangeQuery/expr=changes(a_one[1d]),steps=1-16 391µs ± 2% 542µs ± 1% +38.58% (p=0.000 n=9+8) RangeQuery/expr=changes(a_one[1d]),steps=10-16 452µs ± 2% 617µs ± 2% +36.48% (p=0.000 n=10+10) RangeQuery/expr=changes(a_one[1d]),steps=100-16 1.12ms ± 1% 1.36ms ± 2% +21.58% (p=0.000 n=8+10) RangeQuery/expr=changes(a_one[1d]),steps=1000-16 7.83ms ± 1% 8.94ms ± 1% +14.21% (p=0.000 n=10+10) RangeQuery/expr=changes(a_ten[1d]),steps=1-16 2.98ms ± 0% 3.30ms ± 1% +10.67% (p=0.000 n=9+10) RangeQuery/expr=changes(a_ten[1d]),steps=10-16 3.66ms ± 1% 4.10ms ± 1% +11.82% (p=0.000 n=10+10) RangeQuery/expr=changes(a_ten[1d]),steps=100-16 10.5ms ± 0% 11.8ms ± 1% +12.50% (p=0.000 n=8+10) RangeQuery/expr=changes(a_ten[1d]),steps=1000-16 77.6ms ± 1% 87.4ms ± 1% +12.63% (p=0.000 n=9+9) RangeQuery/expr=changes(a_hundred[1d]),steps=1-16 30.4ms ± 2% 32.8ms ± 1% +8.01% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=10-16 37.1ms ± 2% 40.6ms ± 2% +9.64% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=100-16 105ms ± 1% 117ms ± 1% +11.69% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=1000-16 783ms ± 3% 876ms ± 1% +11.83% (p=0.000 n=9+10) ``` And then runtime v2.39 compared to after this commit: ``` name old time/op new time/op delta RangeQuery/expr=changes(a_one[1d]),steps=1-16 391µs ± 2% 547µs ± 1% +39.84% (p=0.000 n=9+8) RangeQuery/expr=changes(a_one[1d]),steps=10-16 452µs ± 2% 616µs ± 2% +36.15% (p=0.000 n=10+10) RangeQuery/expr=changes(a_one[1d]),steps=100-16 1.12ms ± 1% 1.26ms ± 1% +12.20% (p=0.000 n=8+10) RangeQuery/expr=changes(a_one[1d]),steps=1000-16 7.83ms ± 1% 7.95ms ± 1% +1.59% (p=0.000 n=10+8) RangeQuery/expr=changes(a_ten[1d]),steps=1-16 2.98ms ± 0% 3.38ms ± 2% +13.49% (p=0.000 n=9+10) RangeQuery/expr=changes(a_ten[1d]),steps=10-16 3.66ms ± 1% 4.02ms ± 1% +9.80% (p=0.000 n=10+9) RangeQuery/expr=changes(a_ten[1d]),steps=100-16 10.5ms ± 0% 10.8ms ± 1% +3.08% (p=0.000 n=8+10) RangeQuery/expr=changes(a_ten[1d]),steps=1000-16 77.6ms ± 1% 78.1ms ± 1% +0.58% (p=0.035 n=9+10) RangeQuery/expr=changes(a_hundred[1d]),steps=1-16 30.4ms ± 2% 33.5ms ± 4% +10.18% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=10-16 37.1ms ± 2% 40.0ms ± 1% +7.98% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=100-16 105ms ± 1% 107ms ± 1% +1.92% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=1000-16 783ms ± 3% 775ms ± 1% -1.02% (p=0.019 n=9+9) ``` In summary, the runtime doesn't really improve with this change for queries with just a few steps. For queries with many steps, this commit essentially reinstates the old performance. This is good because the many-step queries are the one that matter most (longest absolute runtime). In terms of allocations, though, this commit doesn't make a dent at all (numbers not shown). The reason is that most of the allocations happen in the sampleRingIterator (in the storage package), which has to be addressed in a separate commit. Signed-off-by: beorn7 <beorn@grafana.com>
2022-10-28 07:58:40 -07:00
l := labels.Labels{}
for {
et, err := p.Next()
if err == io.EOF {
break
}
require.NoError(t, err)
if et == textparse.EntryHistogram || et == textparse.EntrySeries {
p.Metric(&l)
}
switch et {
case textparse.EntryHelp:
metricFamilies++
case textparse.EntryHistogram:
_, parsedTimestamp, h, fh := p.Histogram()
require.Nil(t, h)
promql: Separate `Point` into `FPoint` and `HPoint` In other words: Instead of having a “polymorphous” `Point` that can either contain a float value or a histogram value, use an `FPoint` for floats and an `HPoint` for histograms. This seemingly small change has a _lot_ of repercussions throughout the codebase. The idea here is to avoid the increase in size of `Point` arrays that happened after native histograms had been added. The higher-level data structures (`Sample`, `Series`, etc.) are still “polymorphous”. The same idea could be applied to them, but at each step the trade-offs needed to be evaluated. The idea with this change is to do the minimum necessary to get back to pre-histogram performance for functions that do not touch histograms. Here are comparisons for the `changes` function. The test data doesn't include histograms yet. Ideally, there would be no change in the benchmark result at all. First runtime v2.39 compared to directly prior to this commit: ``` name old time/op new time/op delta RangeQuery/expr=changes(a_one[1d]),steps=1-16 391µs ± 2% 542µs ± 1% +38.58% (p=0.000 n=9+8) RangeQuery/expr=changes(a_one[1d]),steps=10-16 452µs ± 2% 617µs ± 2% +36.48% (p=0.000 n=10+10) RangeQuery/expr=changes(a_one[1d]),steps=100-16 1.12ms ± 1% 1.36ms ± 2% +21.58% (p=0.000 n=8+10) RangeQuery/expr=changes(a_one[1d]),steps=1000-16 7.83ms ± 1% 8.94ms ± 1% +14.21% (p=0.000 n=10+10) RangeQuery/expr=changes(a_ten[1d]),steps=1-16 2.98ms ± 0% 3.30ms ± 1% +10.67% (p=0.000 n=9+10) RangeQuery/expr=changes(a_ten[1d]),steps=10-16 3.66ms ± 1% 4.10ms ± 1% +11.82% (p=0.000 n=10+10) RangeQuery/expr=changes(a_ten[1d]),steps=100-16 10.5ms ± 0% 11.8ms ± 1% +12.50% (p=0.000 n=8+10) RangeQuery/expr=changes(a_ten[1d]),steps=1000-16 77.6ms ± 1% 87.4ms ± 1% +12.63% (p=0.000 n=9+9) RangeQuery/expr=changes(a_hundred[1d]),steps=1-16 30.4ms ± 2% 32.8ms ± 1% +8.01% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=10-16 37.1ms ± 2% 40.6ms ± 2% +9.64% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=100-16 105ms ± 1% 117ms ± 1% +11.69% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=1000-16 783ms ± 3% 876ms ± 1% +11.83% (p=0.000 n=9+10) ``` And then runtime v2.39 compared to after this commit: ``` name old time/op new time/op delta RangeQuery/expr=changes(a_one[1d]),steps=1-16 391µs ± 2% 547µs ± 1% +39.84% (p=0.000 n=9+8) RangeQuery/expr=changes(a_one[1d]),steps=10-16 452µs ± 2% 616µs ± 2% +36.15% (p=0.000 n=10+10) RangeQuery/expr=changes(a_one[1d]),steps=100-16 1.12ms ± 1% 1.26ms ± 1% +12.20% (p=0.000 n=8+10) RangeQuery/expr=changes(a_one[1d]),steps=1000-16 7.83ms ± 1% 7.95ms ± 1% +1.59% (p=0.000 n=10+8) RangeQuery/expr=changes(a_ten[1d]),steps=1-16 2.98ms ± 0% 3.38ms ± 2% +13.49% (p=0.000 n=9+10) RangeQuery/expr=changes(a_ten[1d]),steps=10-16 3.66ms ± 1% 4.02ms ± 1% +9.80% (p=0.000 n=10+9) RangeQuery/expr=changes(a_ten[1d]),steps=100-16 10.5ms ± 0% 10.8ms ± 1% +3.08% (p=0.000 n=8+10) RangeQuery/expr=changes(a_ten[1d]),steps=1000-16 77.6ms ± 1% 78.1ms ± 1% +0.58% (p=0.035 n=9+10) RangeQuery/expr=changes(a_hundred[1d]),steps=1-16 30.4ms ± 2% 33.5ms ± 4% +10.18% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=10-16 37.1ms ± 2% 40.0ms ± 1% +7.98% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=100-16 105ms ± 1% 107ms ± 1% +1.92% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=1000-16 783ms ± 3% 775ms ± 1% -1.02% (p=0.019 n=9+9) ``` In summary, the runtime doesn't really improve with this change for queries with just a few steps. For queries with many steps, this commit essentially reinstates the old performance. This is good because the many-step queries are the one that matter most (longest absolute runtime). In terms of allocations, though, this commit doesn't make a dent at all (numbers not shown). The reason is that most of the allocations happen in the sampleRingIterator (in the storage package), which has to be addressed in a separate commit. Signed-off-by: beorn7 <beorn@grafana.com>
2022-10-28 07:58:40 -07:00
actVec = append(actVec, promql.Sample{
T: *parsedTimestamp,
H: fh,
Metric: l,
})
case textparse.EntrySeries:
promql: Separate `Point` into `FPoint` and `HPoint` In other words: Instead of having a “polymorphous” `Point` that can either contain a float value or a histogram value, use an `FPoint` for floats and an `HPoint` for histograms. This seemingly small change has a _lot_ of repercussions throughout the codebase. The idea here is to avoid the increase in size of `Point` arrays that happened after native histograms had been added. The higher-level data structures (`Sample`, `Series`, etc.) are still “polymorphous”. The same idea could be applied to them, but at each step the trade-offs needed to be evaluated. The idea with this change is to do the minimum necessary to get back to pre-histogram performance for functions that do not touch histograms. Here are comparisons for the `changes` function. The test data doesn't include histograms yet. Ideally, there would be no change in the benchmark result at all. First runtime v2.39 compared to directly prior to this commit: ``` name old time/op new time/op delta RangeQuery/expr=changes(a_one[1d]),steps=1-16 391µs ± 2% 542µs ± 1% +38.58% (p=0.000 n=9+8) RangeQuery/expr=changes(a_one[1d]),steps=10-16 452µs ± 2% 617µs ± 2% +36.48% (p=0.000 n=10+10) RangeQuery/expr=changes(a_one[1d]),steps=100-16 1.12ms ± 1% 1.36ms ± 2% +21.58% (p=0.000 n=8+10) RangeQuery/expr=changes(a_one[1d]),steps=1000-16 7.83ms ± 1% 8.94ms ± 1% +14.21% (p=0.000 n=10+10) RangeQuery/expr=changes(a_ten[1d]),steps=1-16 2.98ms ± 0% 3.30ms ± 1% +10.67% (p=0.000 n=9+10) RangeQuery/expr=changes(a_ten[1d]),steps=10-16 3.66ms ± 1% 4.10ms ± 1% +11.82% (p=0.000 n=10+10) RangeQuery/expr=changes(a_ten[1d]),steps=100-16 10.5ms ± 0% 11.8ms ± 1% +12.50% (p=0.000 n=8+10) RangeQuery/expr=changes(a_ten[1d]),steps=1000-16 77.6ms ± 1% 87.4ms ± 1% +12.63% (p=0.000 n=9+9) RangeQuery/expr=changes(a_hundred[1d]),steps=1-16 30.4ms ± 2% 32.8ms ± 1% +8.01% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=10-16 37.1ms ± 2% 40.6ms ± 2% +9.64% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=100-16 105ms ± 1% 117ms ± 1% +11.69% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=1000-16 783ms ± 3% 876ms ± 1% +11.83% (p=0.000 n=9+10) ``` And then runtime v2.39 compared to after this commit: ``` name old time/op new time/op delta RangeQuery/expr=changes(a_one[1d]),steps=1-16 391µs ± 2% 547µs ± 1% +39.84% (p=0.000 n=9+8) RangeQuery/expr=changes(a_one[1d]),steps=10-16 452µs ± 2% 616µs ± 2% +36.15% (p=0.000 n=10+10) RangeQuery/expr=changes(a_one[1d]),steps=100-16 1.12ms ± 1% 1.26ms ± 1% +12.20% (p=0.000 n=8+10) RangeQuery/expr=changes(a_one[1d]),steps=1000-16 7.83ms ± 1% 7.95ms ± 1% +1.59% (p=0.000 n=10+8) RangeQuery/expr=changes(a_ten[1d]),steps=1-16 2.98ms ± 0% 3.38ms ± 2% +13.49% (p=0.000 n=9+10) RangeQuery/expr=changes(a_ten[1d]),steps=10-16 3.66ms ± 1% 4.02ms ± 1% +9.80% (p=0.000 n=10+9) RangeQuery/expr=changes(a_ten[1d]),steps=100-16 10.5ms ± 0% 10.8ms ± 1% +3.08% (p=0.000 n=8+10) RangeQuery/expr=changes(a_ten[1d]),steps=1000-16 77.6ms ± 1% 78.1ms ± 1% +0.58% (p=0.035 n=9+10) RangeQuery/expr=changes(a_hundred[1d]),steps=1-16 30.4ms ± 2% 33.5ms ± 4% +10.18% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=10-16 37.1ms ± 2% 40.0ms ± 1% +7.98% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=100-16 105ms ± 1% 107ms ± 1% +1.92% (p=0.000 n=10+10) RangeQuery/expr=changes(a_hundred[1d]),steps=1000-16 783ms ± 3% 775ms ± 1% -1.02% (p=0.019 n=9+9) ``` In summary, the runtime doesn't really improve with this change for queries with just a few steps. For queries with many steps, this commit essentially reinstates the old performance. This is good because the many-step queries are the one that matter most (longest absolute runtime). In terms of allocations, though, this commit doesn't make a dent at all (numbers not shown). The reason is that most of the allocations happen in the sampleRingIterator (in the storage package), which has to be addressed in a separate commit. Signed-off-by: beorn7 <beorn@grafana.com>
2022-10-28 07:58:40 -07:00
_, parsedTimestamp, f := p.Series()
actVec = append(actVec, promql.Sample{
T: *parsedTimestamp,
F: f,
Metric: l,
})
}
}
// TODO(codesome): Once PromQL is able to set the CounterResetHint on histograms,
// test it with switching histogram types for metric families.
require.Equal(t, 4, metricFamilies)
require.Equal(t, expVec, actVec)
}