1
0
mirror of https://github.com/golang/go synced 2024-11-18 15:04:44 -07:00
go/internal/event/bench_test.go

155 lines
3.6 KiB
Go
Raw Normal View History

package event_test
telemetry/log: sample reference for benchmarking harness Serves as a reference harness to get benchmarking started. Steps to run benchmark: $ go test -run=^$ -bench=. -count=10 After you grab the output, put the various comparisons into files before.txt and after.txt then edit those result names to have a common name e.g. s/BenchmarkLoggingNoExporter/BenchmarkIt/g s/BenchmarkNoTracingNoMetricsNoLogging/BenchmarkIt/g s/BenchmarkLoggingStdlib/BenchmarkIt/g Now run benchstat: * All compared $ benchstat no_log.txt tellog.txt stdlog.txt name \ time/op no_log.txt tellog.txt stdlog.txt It-8 289ns ± 2% 2780ns ± 3% 5100ns ± 3% name \ alloc/op no_log.txt tellog.txt stdlog.txt It-8 80.0B ± 0% 728.0B ± 0% 568.0B ± 0% name \ allocs/op no_log.txt tellog.txt stdlog.txt It-8 5.00 ± 0% 32.00 ± 0% 28.00 ± 0% * No logging vs telemetry log $ benchstat no_log.txt tellog.txt name old time/op new time/op delta It-8 289ns ± 2% 2780ns ± 3% +862.31% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 728.0B ± 0% +810.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 32.00 ± 0% +540.00% (p=0.000 n=10+10) * No logging vs Standard library "log" $ benchstat no_log.txt stdlog.txt name old time/op new time/op delta It-8 289ns ± 2% 5100ns ± 3% +1665.16% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 568.0B ± 0% +610.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 28.00 ± 0% +460.00% (p=0.000 n=10+10) * telemetry log vs Standard library "log" $ benchstat tellog.txt stdlog.txt name old time/op new time/op delta It-8 2.78µs ± 3% 5.10µs ± 3% +83.43% (p=0.000 n=9+9) name old alloc/op new alloc/op delta It-8 728B ± 0% 568B ± 0% -21.98% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 32.0 ± 0% 28.0 ± 0% -12.50% (p=0.000 n=10+10) Change-Id: I53b15e9da315615278c576f3a60108435417a9f7 Reviewed-on: https://go-review.googlesource.com/c/tools/+/212078 Run-TryBot: Emmanuel Odeke <emm.odeke@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Cottrell <iancottrell@google.com>
2019-12-19 09:57:21 -07:00
import (
"context"
"io/ioutil"
"log"
telemetry/log: sample reference for benchmarking harness Serves as a reference harness to get benchmarking started. Steps to run benchmark: $ go test -run=^$ -bench=. -count=10 After you grab the output, put the various comparisons into files before.txt and after.txt then edit those result names to have a common name e.g. s/BenchmarkLoggingNoExporter/BenchmarkIt/g s/BenchmarkNoTracingNoMetricsNoLogging/BenchmarkIt/g s/BenchmarkLoggingStdlib/BenchmarkIt/g Now run benchstat: * All compared $ benchstat no_log.txt tellog.txt stdlog.txt name \ time/op no_log.txt tellog.txt stdlog.txt It-8 289ns ± 2% 2780ns ± 3% 5100ns ± 3% name \ alloc/op no_log.txt tellog.txt stdlog.txt It-8 80.0B ± 0% 728.0B ± 0% 568.0B ± 0% name \ allocs/op no_log.txt tellog.txt stdlog.txt It-8 5.00 ± 0% 32.00 ± 0% 28.00 ± 0% * No logging vs telemetry log $ benchstat no_log.txt tellog.txt name old time/op new time/op delta It-8 289ns ± 2% 2780ns ± 3% +862.31% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 728.0B ± 0% +810.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 32.00 ± 0% +540.00% (p=0.000 n=10+10) * No logging vs Standard library "log" $ benchstat no_log.txt stdlog.txt name old time/op new time/op delta It-8 289ns ± 2% 5100ns ± 3% +1665.16% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 568.0B ± 0% +610.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 28.00 ± 0% +460.00% (p=0.000 n=10+10) * telemetry log vs Standard library "log" $ benchstat tellog.txt stdlog.txt name old time/op new time/op delta It-8 2.78µs ± 3% 5.10µs ± 3% +83.43% (p=0.000 n=9+9) name old alloc/op new alloc/op delta It-8 728B ± 0% 568B ± 0% -21.98% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 32.0 ± 0% 28.0 ± 0% -12.50% (p=0.000 n=10+10) Change-Id: I53b15e9da315615278c576f3a60108435417a9f7 Reviewed-on: https://go-review.googlesource.com/c/tools/+/212078 Run-TryBot: Emmanuel Odeke <emm.odeke@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Cottrell <iancottrell@google.com>
2019-12-19 09:57:21 -07:00
"testing"
"golang.org/x/tools/internal/event"
"golang.org/x/tools/internal/event/core"
"golang.org/x/tools/internal/event/export"
"golang.org/x/tools/internal/event/keys"
"golang.org/x/tools/internal/event/label"
telemetry/log: sample reference for benchmarking harness Serves as a reference harness to get benchmarking started. Steps to run benchmark: $ go test -run=^$ -bench=. -count=10 After you grab the output, put the various comparisons into files before.txt and after.txt then edit those result names to have a common name e.g. s/BenchmarkLoggingNoExporter/BenchmarkIt/g s/BenchmarkNoTracingNoMetricsNoLogging/BenchmarkIt/g s/BenchmarkLoggingStdlib/BenchmarkIt/g Now run benchstat: * All compared $ benchstat no_log.txt tellog.txt stdlog.txt name \ time/op no_log.txt tellog.txt stdlog.txt It-8 289ns ± 2% 2780ns ± 3% 5100ns ± 3% name \ alloc/op no_log.txt tellog.txt stdlog.txt It-8 80.0B ± 0% 728.0B ± 0% 568.0B ± 0% name \ allocs/op no_log.txt tellog.txt stdlog.txt It-8 5.00 ± 0% 32.00 ± 0% 28.00 ± 0% * No logging vs telemetry log $ benchstat no_log.txt tellog.txt name old time/op new time/op delta It-8 289ns ± 2% 2780ns ± 3% +862.31% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 728.0B ± 0% +810.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 32.00 ± 0% +540.00% (p=0.000 n=10+10) * No logging vs Standard library "log" $ benchstat no_log.txt stdlog.txt name old time/op new time/op delta It-8 289ns ± 2% 5100ns ± 3% +1665.16% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 568.0B ± 0% +610.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 28.00 ± 0% +460.00% (p=0.000 n=10+10) * telemetry log vs Standard library "log" $ benchstat tellog.txt stdlog.txt name old time/op new time/op delta It-8 2.78µs ± 3% 5.10µs ± 3% +83.43% (p=0.000 n=9+9) name old alloc/op new alloc/op delta It-8 728B ± 0% 568B ± 0% -21.98% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 32.0 ± 0% 28.0 ± 0% -12.50% (p=0.000 n=10+10) Change-Id: I53b15e9da315615278c576f3a60108435417a9f7 Reviewed-on: https://go-review.googlesource.com/c/tools/+/212078 Run-TryBot: Emmanuel Odeke <emm.odeke@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Cottrell <iancottrell@google.com>
2019-12-19 09:57:21 -07:00
)
type Hooks struct {
A func(ctx context.Context, a int) (context.Context, func())
B func(ctx context.Context, b string) (context.Context, func())
}
var (
aValue = keys.NewInt("a", "")
bValue = keys.NewString("b", "")
aCount = keys.NewInt64("aCount", "Count of time A is called.")
aStat = keys.NewInt("aValue", "A value.")
bCount = keys.NewInt64("B", "Count of time B is called.")
bLength = keys.NewInt("BLen", "B length.")
Baseline = Hooks{
A: func(ctx context.Context, a int) (context.Context, func()) {
return ctx, func() {}
},
B: func(ctx context.Context, b string) (context.Context, func()) {
return ctx, func() {}
},
telemetry/log: sample reference for benchmarking harness Serves as a reference harness to get benchmarking started. Steps to run benchmark: $ go test -run=^$ -bench=. -count=10 After you grab the output, put the various comparisons into files before.txt and after.txt then edit those result names to have a common name e.g. s/BenchmarkLoggingNoExporter/BenchmarkIt/g s/BenchmarkNoTracingNoMetricsNoLogging/BenchmarkIt/g s/BenchmarkLoggingStdlib/BenchmarkIt/g Now run benchstat: * All compared $ benchstat no_log.txt tellog.txt stdlog.txt name \ time/op no_log.txt tellog.txt stdlog.txt It-8 289ns ± 2% 2780ns ± 3% 5100ns ± 3% name \ alloc/op no_log.txt tellog.txt stdlog.txt It-8 80.0B ± 0% 728.0B ± 0% 568.0B ± 0% name \ allocs/op no_log.txt tellog.txt stdlog.txt It-8 5.00 ± 0% 32.00 ± 0% 28.00 ± 0% * No logging vs telemetry log $ benchstat no_log.txt tellog.txt name old time/op new time/op delta It-8 289ns ± 2% 2780ns ± 3% +862.31% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 728.0B ± 0% +810.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 32.00 ± 0% +540.00% (p=0.000 n=10+10) * No logging vs Standard library "log" $ benchstat no_log.txt stdlog.txt name old time/op new time/op delta It-8 289ns ± 2% 5100ns ± 3% +1665.16% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 568.0B ± 0% +610.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 28.00 ± 0% +460.00% (p=0.000 n=10+10) * telemetry log vs Standard library "log" $ benchstat tellog.txt stdlog.txt name old time/op new time/op delta It-8 2.78µs ± 3% 5.10µs ± 3% +83.43% (p=0.000 n=9+9) name old alloc/op new alloc/op delta It-8 728B ± 0% 568B ± 0% -21.98% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 32.0 ± 0% 28.0 ± 0% -12.50% (p=0.000 n=10+10) Change-Id: I53b15e9da315615278c576f3a60108435417a9f7 Reviewed-on: https://go-review.googlesource.com/c/tools/+/212078 Run-TryBot: Emmanuel Odeke <emm.odeke@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Cottrell <iancottrell@google.com>
2019-12-19 09:57:21 -07:00
}
StdLog = Hooks{
A: func(ctx context.Context, a int) (context.Context, func()) {
log.Printf("A where a=%d", a)
return ctx, func() {}
},
B: func(ctx context.Context, b string) (context.Context, func()) {
log.Printf("B where b=%q", b)
return ctx, func() {}
},
telemetry/log: sample reference for benchmarking harness Serves as a reference harness to get benchmarking started. Steps to run benchmark: $ go test -run=^$ -bench=. -count=10 After you grab the output, put the various comparisons into files before.txt and after.txt then edit those result names to have a common name e.g. s/BenchmarkLoggingNoExporter/BenchmarkIt/g s/BenchmarkNoTracingNoMetricsNoLogging/BenchmarkIt/g s/BenchmarkLoggingStdlib/BenchmarkIt/g Now run benchstat: * All compared $ benchstat no_log.txt tellog.txt stdlog.txt name \ time/op no_log.txt tellog.txt stdlog.txt It-8 289ns ± 2% 2780ns ± 3% 5100ns ± 3% name \ alloc/op no_log.txt tellog.txt stdlog.txt It-8 80.0B ± 0% 728.0B ± 0% 568.0B ± 0% name \ allocs/op no_log.txt tellog.txt stdlog.txt It-8 5.00 ± 0% 32.00 ± 0% 28.00 ± 0% * No logging vs telemetry log $ benchstat no_log.txt tellog.txt name old time/op new time/op delta It-8 289ns ± 2% 2780ns ± 3% +862.31% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 728.0B ± 0% +810.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 32.00 ± 0% +540.00% (p=0.000 n=10+10) * No logging vs Standard library "log" $ benchstat no_log.txt stdlog.txt name old time/op new time/op delta It-8 289ns ± 2% 5100ns ± 3% +1665.16% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 568.0B ± 0% +610.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 28.00 ± 0% +460.00% (p=0.000 n=10+10) * telemetry log vs Standard library "log" $ benchstat tellog.txt stdlog.txt name old time/op new time/op delta It-8 2.78µs ± 3% 5.10µs ± 3% +83.43% (p=0.000 n=9+9) name old alloc/op new alloc/op delta It-8 728B ± 0% 568B ± 0% -21.98% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 32.0 ± 0% 28.0 ± 0% -12.50% (p=0.000 n=10+10) Change-Id: I53b15e9da315615278c576f3a60108435417a9f7 Reviewed-on: https://go-review.googlesource.com/c/tools/+/212078 Run-TryBot: Emmanuel Odeke <emm.odeke@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Cottrell <iancottrell@google.com>
2019-12-19 09:57:21 -07:00
}
Log = Hooks{
A: func(ctx context.Context, a int) (context.Context, func()) {
core.Log1(ctx, "A", aValue.Of(a))
return ctx, func() {}
},
B: func(ctx context.Context, b string) (context.Context, func()) {
core.Log1(ctx, "B", bValue.Of(b))
return ctx, func() {}
},
telemetry/log: sample reference for benchmarking harness Serves as a reference harness to get benchmarking started. Steps to run benchmark: $ go test -run=^$ -bench=. -count=10 After you grab the output, put the various comparisons into files before.txt and after.txt then edit those result names to have a common name e.g. s/BenchmarkLoggingNoExporter/BenchmarkIt/g s/BenchmarkNoTracingNoMetricsNoLogging/BenchmarkIt/g s/BenchmarkLoggingStdlib/BenchmarkIt/g Now run benchstat: * All compared $ benchstat no_log.txt tellog.txt stdlog.txt name \ time/op no_log.txt tellog.txt stdlog.txt It-8 289ns ± 2% 2780ns ± 3% 5100ns ± 3% name \ alloc/op no_log.txt tellog.txt stdlog.txt It-8 80.0B ± 0% 728.0B ± 0% 568.0B ± 0% name \ allocs/op no_log.txt tellog.txt stdlog.txt It-8 5.00 ± 0% 32.00 ± 0% 28.00 ± 0% * No logging vs telemetry log $ benchstat no_log.txt tellog.txt name old time/op new time/op delta It-8 289ns ± 2% 2780ns ± 3% +862.31% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 728.0B ± 0% +810.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 32.00 ± 0% +540.00% (p=0.000 n=10+10) * No logging vs Standard library "log" $ benchstat no_log.txt stdlog.txt name old time/op new time/op delta It-8 289ns ± 2% 5100ns ± 3% +1665.16% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 568.0B ± 0% +610.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 28.00 ± 0% +460.00% (p=0.000 n=10+10) * telemetry log vs Standard library "log" $ benchstat tellog.txt stdlog.txt name old time/op new time/op delta It-8 2.78µs ± 3% 5.10µs ± 3% +83.43% (p=0.000 n=9+9) name old alloc/op new alloc/op delta It-8 728B ± 0% 568B ± 0% -21.98% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 32.0 ± 0% 28.0 ± 0% -12.50% (p=0.000 n=10+10) Change-Id: I53b15e9da315615278c576f3a60108435417a9f7 Reviewed-on: https://go-review.googlesource.com/c/tools/+/212078 Run-TryBot: Emmanuel Odeke <emm.odeke@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Cottrell <iancottrell@google.com>
2019-12-19 09:57:21 -07:00
}
Trace = Hooks{
A: func(ctx context.Context, a int) (context.Context, func()) {
return core.Start1(ctx, "A", aValue.Of(a))
},
B: func(ctx context.Context, b string) (context.Context, func()) {
return core.Start1(ctx, "B", bValue.Of(b))
},
telemetry/log: sample reference for benchmarking harness Serves as a reference harness to get benchmarking started. Steps to run benchmark: $ go test -run=^$ -bench=. -count=10 After you grab the output, put the various comparisons into files before.txt and after.txt then edit those result names to have a common name e.g. s/BenchmarkLoggingNoExporter/BenchmarkIt/g s/BenchmarkNoTracingNoMetricsNoLogging/BenchmarkIt/g s/BenchmarkLoggingStdlib/BenchmarkIt/g Now run benchstat: * All compared $ benchstat no_log.txt tellog.txt stdlog.txt name \ time/op no_log.txt tellog.txt stdlog.txt It-8 289ns ± 2% 2780ns ± 3% 5100ns ± 3% name \ alloc/op no_log.txt tellog.txt stdlog.txt It-8 80.0B ± 0% 728.0B ± 0% 568.0B ± 0% name \ allocs/op no_log.txt tellog.txt stdlog.txt It-8 5.00 ± 0% 32.00 ± 0% 28.00 ± 0% * No logging vs telemetry log $ benchstat no_log.txt tellog.txt name old time/op new time/op delta It-8 289ns ± 2% 2780ns ± 3% +862.31% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 728.0B ± 0% +810.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 32.00 ± 0% +540.00% (p=0.000 n=10+10) * No logging vs Standard library "log" $ benchstat no_log.txt stdlog.txt name old time/op new time/op delta It-8 289ns ± 2% 5100ns ± 3% +1665.16% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 568.0B ± 0% +610.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 28.00 ± 0% +460.00% (p=0.000 n=10+10) * telemetry log vs Standard library "log" $ benchstat tellog.txt stdlog.txt name old time/op new time/op delta It-8 2.78µs ± 3% 5.10µs ± 3% +83.43% (p=0.000 n=9+9) name old alloc/op new alloc/op delta It-8 728B ± 0% 568B ± 0% -21.98% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 32.0 ± 0% 28.0 ± 0% -12.50% (p=0.000 n=10+10) Change-Id: I53b15e9da315615278c576f3a60108435417a9f7 Reviewed-on: https://go-review.googlesource.com/c/tools/+/212078 Run-TryBot: Emmanuel Odeke <emm.odeke@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Cottrell <iancottrell@google.com>
2019-12-19 09:57:21 -07:00
}
Stats = Hooks{
A: func(ctx context.Context, a int) (context.Context, func()) {
core.Metric1(ctx, aStat.Of(a))
core.Metric1(ctx, aCount.Of(1))
return ctx, func() {}
},
B: func(ctx context.Context, b string) (context.Context, func()) {
core.Metric1(ctx, bLength.Of(len(b)))
core.Metric1(ctx, bCount.Of(1))
return ctx, func() {}
},
}
initialList = []int{0, 1, 22, 333, 4444, 55555, 666666, 7777777}
stringList = []string{
"A value",
"Some other value",
"A nice longer value but not too long",
"V",
"",
"ı",
"prime count of values",
}
)
type namedBenchmark struct {
name string
test func(*testing.B)
}
func Benchmark(b *testing.B) {
b.Run("Baseline", Baseline.runBenchmark)
b.Run("StdLog", StdLog.runBenchmark)
benchmarks := []namedBenchmark{
{"Log", Log.runBenchmark},
{"Trace", Trace.runBenchmark},
{"Stats", Stats.runBenchmark},
}
event.SetExporter(nil)
for _, t := range benchmarks {
b.Run(t.name+"NoExporter", t.test)
}
event.SetExporter(noopExporter)
for _, t := range benchmarks {
b.Run(t.name+"Noop", t.test)
}
event.SetExporter(export.Spans(export.LogWriter(ioutil.Discard, false)))
for _, t := range benchmarks {
b.Run(t.name, t.test)
}
telemetry/log: sample reference for benchmarking harness Serves as a reference harness to get benchmarking started. Steps to run benchmark: $ go test -run=^$ -bench=. -count=10 After you grab the output, put the various comparisons into files before.txt and after.txt then edit those result names to have a common name e.g. s/BenchmarkLoggingNoExporter/BenchmarkIt/g s/BenchmarkNoTracingNoMetricsNoLogging/BenchmarkIt/g s/BenchmarkLoggingStdlib/BenchmarkIt/g Now run benchstat: * All compared $ benchstat no_log.txt tellog.txt stdlog.txt name \ time/op no_log.txt tellog.txt stdlog.txt It-8 289ns ± 2% 2780ns ± 3% 5100ns ± 3% name \ alloc/op no_log.txt tellog.txt stdlog.txt It-8 80.0B ± 0% 728.0B ± 0% 568.0B ± 0% name \ allocs/op no_log.txt tellog.txt stdlog.txt It-8 5.00 ± 0% 32.00 ± 0% 28.00 ± 0% * No logging vs telemetry log $ benchstat no_log.txt tellog.txt name old time/op new time/op delta It-8 289ns ± 2% 2780ns ± 3% +862.31% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 728.0B ± 0% +810.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 32.00 ± 0% +540.00% (p=0.000 n=10+10) * No logging vs Standard library "log" $ benchstat no_log.txt stdlog.txt name old time/op new time/op delta It-8 289ns ± 2% 5100ns ± 3% +1665.16% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 568.0B ± 0% +610.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 28.00 ± 0% +460.00% (p=0.000 n=10+10) * telemetry log vs Standard library "log" $ benchstat tellog.txt stdlog.txt name old time/op new time/op delta It-8 2.78µs ± 3% 5.10µs ± 3% +83.43% (p=0.000 n=9+9) name old alloc/op new alloc/op delta It-8 728B ± 0% 568B ± 0% -21.98% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 32.0 ± 0% 28.0 ± 0% -12.50% (p=0.000 n=10+10) Change-Id: I53b15e9da315615278c576f3a60108435417a9f7 Reviewed-on: https://go-review.googlesource.com/c/tools/+/212078 Run-TryBot: Emmanuel Odeke <emm.odeke@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Cottrell <iancottrell@google.com>
2019-12-19 09:57:21 -07:00
}
func A(ctx context.Context, hooks Hooks, a int) int {
ctx, done := hooks.A(ctx, a)
defer done()
return B(ctx, hooks, a, stringList[a%len(stringList)])
}
func B(ctx context.Context, hooks Hooks, a int, b string) int {
_, done := hooks.B(ctx, b)
defer done()
return a + len(b)
}
func (hooks Hooks) runBenchmark(b *testing.B) {
ctx := context.Background()
b.ReportAllocs()
b.ResetTimer()
var acc int
for i := 0; i < b.N; i++ {
for _, value := range initialList {
acc += A(ctx, hooks, value)
}
telemetry/log: sample reference for benchmarking harness Serves as a reference harness to get benchmarking started. Steps to run benchmark: $ go test -run=^$ -bench=. -count=10 After you grab the output, put the various comparisons into files before.txt and after.txt then edit those result names to have a common name e.g. s/BenchmarkLoggingNoExporter/BenchmarkIt/g s/BenchmarkNoTracingNoMetricsNoLogging/BenchmarkIt/g s/BenchmarkLoggingStdlib/BenchmarkIt/g Now run benchstat: * All compared $ benchstat no_log.txt tellog.txt stdlog.txt name \ time/op no_log.txt tellog.txt stdlog.txt It-8 289ns ± 2% 2780ns ± 3% 5100ns ± 3% name \ alloc/op no_log.txt tellog.txt stdlog.txt It-8 80.0B ± 0% 728.0B ± 0% 568.0B ± 0% name \ allocs/op no_log.txt tellog.txt stdlog.txt It-8 5.00 ± 0% 32.00 ± 0% 28.00 ± 0% * No logging vs telemetry log $ benchstat no_log.txt tellog.txt name old time/op new time/op delta It-8 289ns ± 2% 2780ns ± 3% +862.31% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 728.0B ± 0% +810.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 32.00 ± 0% +540.00% (p=0.000 n=10+10) * No logging vs Standard library "log" $ benchstat no_log.txt stdlog.txt name old time/op new time/op delta It-8 289ns ± 2% 5100ns ± 3% +1665.16% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 568.0B ± 0% +610.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 28.00 ± 0% +460.00% (p=0.000 n=10+10) * telemetry log vs Standard library "log" $ benchstat tellog.txt stdlog.txt name old time/op new time/op delta It-8 2.78µs ± 3% 5.10µs ± 3% +83.43% (p=0.000 n=9+9) name old alloc/op new alloc/op delta It-8 728B ± 0% 568B ± 0% -21.98% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 32.0 ± 0% 28.0 ± 0% -12.50% (p=0.000 n=10+10) Change-Id: I53b15e9da315615278c576f3a60108435417a9f7 Reviewed-on: https://go-review.googlesource.com/c/tools/+/212078 Run-TryBot: Emmanuel Odeke <emm.odeke@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Cottrell <iancottrell@google.com>
2019-12-19 09:57:21 -07:00
}
}
func init() {
log.SetOutput(ioutil.Discard)
telemetry/log: sample reference for benchmarking harness Serves as a reference harness to get benchmarking started. Steps to run benchmark: $ go test -run=^$ -bench=. -count=10 After you grab the output, put the various comparisons into files before.txt and after.txt then edit those result names to have a common name e.g. s/BenchmarkLoggingNoExporter/BenchmarkIt/g s/BenchmarkNoTracingNoMetricsNoLogging/BenchmarkIt/g s/BenchmarkLoggingStdlib/BenchmarkIt/g Now run benchstat: * All compared $ benchstat no_log.txt tellog.txt stdlog.txt name \ time/op no_log.txt tellog.txt stdlog.txt It-8 289ns ± 2% 2780ns ± 3% 5100ns ± 3% name \ alloc/op no_log.txt tellog.txt stdlog.txt It-8 80.0B ± 0% 728.0B ± 0% 568.0B ± 0% name \ allocs/op no_log.txt tellog.txt stdlog.txt It-8 5.00 ± 0% 32.00 ± 0% 28.00 ± 0% * No logging vs telemetry log $ benchstat no_log.txt tellog.txt name old time/op new time/op delta It-8 289ns ± 2% 2780ns ± 3% +862.31% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 728.0B ± 0% +810.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 32.00 ± 0% +540.00% (p=0.000 n=10+10) * No logging vs Standard library "log" $ benchstat no_log.txt stdlog.txt name old time/op new time/op delta It-8 289ns ± 2% 5100ns ± 3% +1665.16% (p=0.000 n=10+9) name old alloc/op new alloc/op delta It-8 80.0B ± 0% 568.0B ± 0% +610.00% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 5.00 ± 0% 28.00 ± 0% +460.00% (p=0.000 n=10+10) * telemetry log vs Standard library "log" $ benchstat tellog.txt stdlog.txt name old time/op new time/op delta It-8 2.78µs ± 3% 5.10µs ± 3% +83.43% (p=0.000 n=9+9) name old alloc/op new alloc/op delta It-8 728B ± 0% 568B ± 0% -21.98% (p=0.000 n=10+10) name old allocs/op new allocs/op delta It-8 32.0 ± 0% 28.0 ± 0% -12.50% (p=0.000 n=10+10) Change-Id: I53b15e9da315615278c576f3a60108435417a9f7 Reviewed-on: https://go-review.googlesource.com/c/tools/+/212078 Run-TryBot: Emmanuel Odeke <emm.odeke@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Cottrell <iancottrell@google.com>
2019-12-19 09:57:21 -07:00
}
func noopExporter(ctx context.Context, ev core.Event, lm label.Map) context.Context {
return ctx
}