Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
You, latency and profiling @ GolangUK 2017
Search
Filippo Valsorda
August 17, 2017
Programming
0
1.2k
You, latency and profiling @ GolangUK 2017
Filippo Valsorda
August 17, 2017
Tweet
Share
More Decks by Filippo Valsorda
See All by Filippo Valsorda
Asynchronous networking @ GopherCon 2018
filosottile
3
2.1k
Le note cifrate di Antonio Marzi
filosottile
1
400
Why cgo is slow @ CapitalGo 2018
filosottile
2
4.5k
Squeezing a key through a carry bit @ 34c3
filosottile
0
1.7k
Calling Rust from Go, without cgo @ GothamGo 2017
filosottile
1
2.7k
Encrypting the Internet with Go @ GopherCon 2017
filosottile
9
2.6k
You, latency and profiling @ GopherCon India 2017
filosottile
13
4k
TLS 1.3 @ 33c3
filosottile
4
6.9k
Stealing Bitcoin With Math - HOPE XI
filosottile
13
30k
Other Decks in Programming
See All in Programming
Goで作る、開発・CI環境
sin392
0
240
GitHub Copilot and GitHub Codespaces Hands-on
ymd65536
2
150
AI駆動のマルチエージェントによる業務フロー自動化の設計と実践
h_okkah
0
170
MDN Web Docs に日本語翻訳でコントリビュートしたくなる
ohmori_yusuke
1
130
『自分のデータだけ見せたい!』を叶える──Laravel × Casbin で複雑権限をスッキリ解きほぐす 25 分
akitotsukahara
2
640
プロダクト志向なエンジニアがもう一歩先の価値を目指すために意識したこと
nealle
0
130
Systèmes distribués, pour le meilleur et pour le pire - BreizhCamp 2025 - Conférence
slecache
0
120
AI時代のソフトウェア開発を考える(2025/07版) / Agentic Software Engineering Findy 2025-07 Edition
twada
PRO
91
31k
Team operations that are not burdened by SRE
kazatohiei
1
310
今ならAmazon ECSのサービス間通信をどう選ぶか / Selection of ECS Interservice Communication 2025
tkikuc
21
4k
#kanrk08 / 公開版 PicoRubyとマイコンでの自作トレーニング計測装置を用いたワークアウトの理想と現実
bash0c7
1
780
なぜ適用するか、移行して理解するClean Architecture 〜構造を超えて設計を継承する〜 / Why Apply, Migrate and Understand Clean Architecture - Inherit Design Beyond Structure
seike460
PRO
3
780
Featured
See All Featured
Rebuilding a faster, lazier Slack
samanthasiow
83
9.1k
Chrome DevTools: State of the Union 2024 - Debugging React & Beyond
addyosmani
7
740
A Modern Web Designer's Workflow
chriscoyier
695
190k
Designing for humans not robots
tammielis
253
25k
Cheating the UX When There Is Nothing More to Optimize - PixelPioneers
stephaniewalter
281
13k
Code Review Best Practice
trishagee
69
19k
Code Reviewing Like a Champion
maltzj
524
40k
Mobile First: as difficult as doing things right
swwweet
223
9.7k
Exploring the Power of Turbo Streams & Action Cable | RailsConf2023
kevinliebholz
34
5.9k
The MySQL Ecosystem @ GitHub 2015
samlambert
251
13k
4 Signs Your Business is Dying
shpigford
184
22k
CSS Pre-Processors: Stylus, Less & Sass
bermonpainter
357
30k
Transcript
You, latency and profiling Filippo Valsorda — @FiloSo/le
What is fast?
A fast hash… … can hash many MB per second
A fast regex… … can process many MB/s
A fast database… … can store many GB per second
A fast database… … can store many GB per second
… or replies to queries in a few ms
A fast website… … can handle many requests … or
loads in a few milliseconds
A fast API… … can handle many clients … or
answers in a few milliseconds
Fast is… Throughput and Latency
CPU profiling -cpuprofile /debug/pprof/profile
SIGPROF CPU profiling
runtime/proc.go
h>p.Handler 1 Running on CPU h>p.Handler 2 I/O I/O h>p.Handler
3 I/O h>p.Handler 4 I/O CPU profiling
Running on CPU I/O I/O I/O I/O h>p.Handler 1 h>p.Handler
2 h>p.Handler 3 h>p.Handler 4 CPU profiling
I/O Observed latency h>p.Handler 1 CPU profiling
func Write(data []byte) { for i := 0; i <
50; i++ { tmpfile, _ := ioutil.TempFile("", "ex") defer os.Remove(tmpfile.Name()) _, err = tmpfile.Write(data) tmpfile.Close() } } func Hash(data []byte) { for i := 0; i < 50; i++ { sha256.Sum256(data) } }
$ time curl http://127.0.0.1:12345/hash-and-write 9.831 total go tool pprof -web
slowserver cpu.pprof
None
$ time curl http://127.0.0.1:12345/hash-and-write 9.831 total $ time
curl http://127.0.0.1:12345/write-no-hash 7.692 total
Latency is not all about CPU
CPU profiling is for throughput The tracer is for latency
$ time curl http://127.0.0.1:12345/write-no-hash 7.692 total go tool trace -pprof=syscall
trace.out > syscall.pprof go tool pprof -web syscall.pprof
None
The tracer Detailed nanosecond-level log of execuOon events: • gorouOne
scheduling, blocking • syscalls, network, I/O • garbage collecOon
•log files are BIG •25% overhead (aWer Go 1.7) •symbols
are embedded (Go 1.7) The tracer
•Full event context •GorouOne metadata and history •Discrete events, not
samples Trace vs. profile
CPU profiling Total: 20 samples
Tracing events
import "runtime/trace" import "net/http/pprof" /debug/pprof/trace?seconds=5 The tracer
None
go tool trace -pprof=TYPE trace.out Supported profile types are: -
net: network blocking profile - sync: synchronization blocking profile - syscall: syscall blocking profile - sched: scheduler latency profile
go tool trace -pprof=syscall trace.out > syscall.pprof go tool pprof
-web syscall.pprof
go tool trace -pprof=syscall trace.out > syscall.pprof go-torch -b syscall.pprof
&& open torch.svg
None
None
-pprof=syscall func Write(data []byte) { for i := 0; i
< 50; i++ { tmpfile, _ := ioutil.TempFile("", "ex") defer os.Remove(tmpfile.Name()) _, err = tmpfile.Write(data) tmpfile.Close() } }
-pprof=sync // Block waits on a channel func Block() {
<-time.After(1 * time.Second) }
-pprof=sync
-pprof=sync
-pprof=net func Download() []byte { r, err := http.Get("https://golang.org/doc/go1.8") body,
err := ioutil.ReadAll(r.Body) r.Body.Close() return body }
-pprof=net
-pprof=net
-pprof=net
None
WriIng our own profile prof := make(map[uint64]pprof.Record) for _, ev
:= range events { if ev.Type != trace.EvGoBlockNet { continue } rec := prof[ev.StkID] rec.Stk = ev.Stk rec.N++ rec.Time += ev.Link.Ts - ev.Ts prof[ev.StkID] = rec } pprof.BuildProfile(prof).Write(os.Stdout)
var childG = make(map[uint64]struct{}) var lastGLen int for { for
_, ev := range events { if ev.Type != trace.EvGoCreate { continue } if _, ok := childG[ev.G]; !ok && !filterStack(ev.Stk, re) { continue } childG[ev.Args[0]] = struct{}{} } if len(childG) == lastGLen { break } lastGLen = len(childG) }
tracefocus tracefocus -filter=ServeHTTP trace.out > net.pprof github.com/FiloSottile/tracetools
None
None
Build more tools! • Focus on gorouOne number • Aggregate
all blocking types • VisualizaOons • …
tracehist tracehist -filter=Close trace.out github.com/FiloSottile/tracetools
None
Build more tools! Profile latency!
QuesIons? Filippo Valsorda @FiloSo/le hi@filippo.io