Go, for Distributed Systems

Russ Cox

Google

About the Talk

I gave variants of this talk three times in 2013, once at SOSP's Programming Languages and Operating Systems (PLOS) workshop, once at MIT Lincoln Lab's annual Software Engineering Symposium, and once at Twitter's Cambridge, Massachusetts office.

The talk assumes an audience familiar with the basic problems of building distributed systems. It presents Go's approach to solving some of those problems.

2

Go

Go is an open source programming language that makes it easy to build simple, reliable, and efficient software.

3

History

Design began in late 2007.

Became open source in November 2009.

Developed entirely in the open; very active community.
Language stable as of Go 1, early 2012.

4

Motivation

Started as an answer to software problems at Google:

5

Go

A simple but powerful and fun language.

For more background on design:

6

This Talk

Engineering

Interfaces

Concurrency

7

Engineering

8

Engineering: Imports

package main

import "fmt"

func main() {
    fmt.Printf("hello, world\n")
}

import "fmt" guaranteed to read exactly one file.

9

Engineering: Imports

$ go get github.com/golang/glog
package main

import (
    "flag"

    "github.com/golang/glog"
)

func main() {
    flag.Set("logtostderr", "true")
    glog.Infof("hello, world")
}

Still guaranteed to read exactly one file.

Import name space is decentralized.

10

Engineering: Program Rewrites

$ gofmt -r 'glog.Infof -> glog.Errorf' hello1.go
package main

import (
    "flag"
    "github.com/golang/glog"
)

func main() {
    flag.Set("logtostderr", "true")
    glog.Errorf("hello, world")
}
$
11

Engineering: Garbage Collection

In C and C++, too much programming and API design is about memory management.

Go has garbage collection, only.

Fundamental for interfaces: memory management details do not bifurcate otherwise-similar APIs.

Fundamental for concurrency: too hard to track ownership otherwise.

Of course, adds cost, latency, complexity in run time system.

12

Engineering: Garbage Collection

Experience with Java: Uncontrollable cost, too much tuning.

Go lets you limit allocation by controlling memory layout.

Examples: 

type Ring struct {
    R, W int
    Data [512]byte
}

type Point struct {
    X, Y int
}

type Rectangle struct {
    Min, Max Point
}
13

Engineering: Garbage Collection

Garbage collector implementation remains an active area of work and research.

Design decision: Interior pointers are allowed, as are foreign pointers.

Current design: parallel mark-and-sweep.
With care to use memory wisely, works well in production.

14

Interfaces

15

Interfaces

An interface defines a set of methods. 

package io

type Writer interface {
    Write(data []byte) (n int, err error)
}
16

Interfaces

A type implements the interface by implementing the methods. 

package bytes

type Buffer struct {
    ...
}

func (b *Buffer) Write(data []byte) (n int, err error) {
    ...
}
17

Interfaces

An implementation of an interface can be assigned to a variable of that interface type. 

package fmt

func Fprintf(w io.Writer, format string, args ...interface{})
18

Interfaces

// +build ignore,OMIT

package main

import (
	"bytes"
	"fmt"
	"io"
	"os"
)

var _ = io.Copy

func main() {

    b := new(bytes.Buffer)
    var w io.Writer
    w = b
    fmt.Fprintf(w, "hello, %s\n", "world")
    os.Stdout.Write(b.Bytes())

}
19

Interfaces

Reader is the obvious counterpart. 

package io

type Reader interface {
    Read(data []byte) (n int, err error)
}

func Copy(dst Writer, src Reader) (n int64, err error)
20

Interfaces

// +build ignore,OMIT

package main

import (
	"bytes"
	"fmt"
	"io"
	"os"
)

func main() {

    b := new(bytes.Buffer)
    fmt.Fprintf(b, "hello, %s\n", "world")
    io.Copy(os.Stdout, b)

}
21

Interfaces

Reader and Writer turn out to be very useful.

Adapters 

package io

func MultiWriter(writers ...Writer) Writer
    MultiWriter creates a writer that duplicates its writes to all the
    provided writers, similar to the Unix tee(1) command.

Chaining 

package gzip // compress/gzip

func NewWriter(w io.Writer) *Writer

Also: buffered writers, encrypted writers, limited writers, HTTP responses.

22

Interfaces

Networking: 

package net

type Conn interface {
    Read(b []byte) (n int, err error)
    Write(b []byte) (n int, err error)
    Close() error

    LocalAddr() Addr
    RemoteAddr() Addr

    SetDeadline(t time.Time) error
    SetReadDeadline(t time.Time) error
    SetWriteDeadline(t time.Time) error
}

func Dial(network, address string) (Conn, error)
23

Interfaces

Networking example: 

// +build ignore,OMIT

package main

import (
	"bufio"
	"io"
	"log"
	"net"
	"os"
	"os/exec"
)

func main() {
	if len(os.Args) > 1 && os.Args[1] == "serve" {
		serve()
	}
	finger()
}

func finger() {

    c, err := net.Dial("tcp", "localhost:finger")
    if err != nil {
        log.Fatal(err)
    }
    io.WriteString(c, "rsc\n")
    io.Copy(os.Stdout, c)

}

func serve() {
	l, err := net.Listen("tcp", "localhost:finger")
	if err != nil {
		log.Fatal(err)
	}
	for {
		c, err := l.Accept()
		if err != nil {
			log.Fatal(err)
		}
		go serveConn(c)
	}
}

func serveConn(c net.Conn) {
	defer c.Close()

	b := bufio.NewReader(c)
	l, err := b.ReadString('\n')
	if err != nil {
		return
	}

	cmd := exec.Command("finger", l[:len(l)-1])
	cmd.Stdout = c
	cmd.Stderr = c
	cmd.Run()
}
24

Interfaces

Networking client as adapter function: 

package smtp

func NewClient(conn net.Conn, host string) (*Client, error)

Other implementations of net.Conn: testing, SSL, ...

25

Interface Lessons

Key advantages:

The source of all generality in the Go language.

26

Concurrency

27

Concurrency vs Parallelism

Concurrency is about dealing with lots of things at once.

Parallelism is about doing lots of things at once.

Concurrency is about structure, parallelism is about execution.

Concurrency provides a way to structure a solution to solve a problem that may be parallelizable (or not).

28

Concurrency vs Parallelism

Concurrent: mouse, keyboard, display, and disk drivers in operating system.

Parallel: vector dot product, matrix multiply.

Concurrency can enable parallelism but is useful on its own: modern programs must deal with many things at once.

29

Concurrency

Go provides two important concepts:

A goroutine is a thread of control within the program, with its own local variables and stack. Cheap, easy to create.

A channel carries typed messages between goroutines.

30

Concurrency

package main

import "fmt"

func main() {
    c := make(chan string)
    go func() {
        c <- "Hello"
        c <- "World"
    }()
    fmt.Println(<-c, <-c)
}
31

Concurrency: CSP

Channels adopted from Hoare's Communicating Sequential Processes.

Go enables simple, safe concurrent programming.
It doesn't forbid bad programming.

Caveat: not purely memory safe; sharing is legal.
Passing a pointer over a channel is idiomatic.

Experience shows this is practical.

32

Concurrency

Sequential network address resolution, given a work list: 

// +build ignore,OMIT

package main

import (
	"fmt"
	"math/rand"
	"time"
)

func lookup() {

    for _, w := range worklist {
        w.addrs, w.err = LookupHost(w.host)
    }

}

func main() {
	rand.Seed(time.Now().UnixNano())

	t0 := time.Now()
	lookup()

	fmt.Printf("\n")
	for _, w := range worklist {
		if w.err != nil {
			fmt.Printf("%s: error: %v\n", w.host, w.err)
			continue
		}
		fmt.Printf("%s: %v\n", w.host, w.addrs)
	}
	fmt.Printf("total lookup time: %.3f seconds\n", time.Since(t0).Seconds())
}

var worklist = []*Work{
	{host: "fast.com"},
	{host: "slow.com"},
	{host: "fast.missing.com"},
	{host: "slow.missing.com"},
}

type Work struct {
	host  string
	addrs []string
	err   error
}

func LookupHost(name string) (addrs []string, err error) {
	t0 := time.Now()
	defer func() {
		fmt.Printf("lookup %s: %.3f seconds\n", name, time.Since(t0).Seconds())
	}()
	h := hosts[name]
	if h == nil {
		h = failure
	}
	return h(name)
}

type resolver func(string) ([]string, error)

var hosts = map[string]resolver{
	"fast.com":         delay(10*time.Millisecond, fixedAddrs("10.0.0.1")),
	"slow.com":         delay(2*time.Second, fixedAddrs("10.0.0.4")),
	"fast.missing.com": delay(10*time.Millisecond, failure),
	"slow.missing.com": delay(2*time.Second, failure),
}

func fixedAddrs(addrs ...string) resolver {
	return func(string) ([]string, error) {
		return addrs, nil
	}
}

func delay(d time.Duration, f resolver) resolver {
	return func(name string) ([]string, error) {
		time.Sleep(d/2 + time.Duration(rand.Int63n(int64(d/2))))
		return f(name)
	}
}

func failure(name string) ([]string, error) {
	return nil, fmt.Errorf("unknown host %v", name)
}
33

Concurrency

Parallel network address resolution, given a work list: 

// +build ignore,OMIT

package main

import (
	"fmt"
	"math/rand"
	"time"
)

func lookup() {

    done := make(chan bool, len(worklist))

    for _, w := range worklist {
        go func(w *Work) {
            w.addrs, w.err = LookupHost(w.host)
            done <- true
        }(w)
    }

    for i := 0; i < len(worklist); i++ {
        <-done
    }

}

func main() {
	rand.Seed(time.Now().UnixNano())

	t0 := time.Now()
	lookup()

	fmt.Printf("\n")
	for _, w := range worklist {
		if w.err != nil {
			fmt.Printf("%s: error: %v\n", w.host, w.err)
			continue
		}
		fmt.Printf("%s: %v\n", w.host, w.addrs)
	}
	fmt.Printf("total lookup time: %.3f seconds\n", time.Since(t0).Seconds())
}

var worklist = []*Work{
	{host: "fast.com"},
	{host: "slow.com"},
	{host: "fast.missing.com"},
	{host: "slow.missing.com"},
}

type Work struct {
	host  string
	addrs []string
	err   error
}

func LookupHost(name string) (addrs []string, err error) {
	t0 := time.Now()
	defer func() {
		fmt.Printf("lookup %s: %.3f seconds\n", name, time.Since(t0).Seconds())
	}()
	h := hosts[name]
	if h == nil {
		h = failure
	}
	return h(name)
}

type resolver func(string) ([]string, error)

var hosts = map[string]resolver{
	"fast.com":         delay(10*time.Millisecond, fixedAddrs("10.0.0.1")),
	"slow.com":         delay(2*time.Second, fixedAddrs("10.0.0.4")),
	"fast.missing.com": delay(10*time.Millisecond, failure),
	"slow.missing.com": delay(2*time.Second, failure),
}

func fixedAddrs(addrs ...string) resolver {
	return func(string) ([]string, error) {
		return addrs, nil
	}
}

func delay(d time.Duration, f resolver) resolver {
	return func(name string) ([]string, error) {
		time.Sleep(d/2 + time.Duration(rand.Int63n(int64(d/2))))
		return f(name)
	}
}

func failure(name string) ([]string, error) {
	return nil, fmt.Errorf("unknown host %v", name)
}
34

Concurrency

Aside: can abstract this pattern. 

// +build ignore,OMIT

package main

import (
	"fmt"
	"math/rand"
	"sync"
	"time"
)

func lookup() {

    var group sync.WaitGroup

    for _, w := range worklist {
        group.Add(1)
        go func(w *Work) {
            w.addrs, w.err = LookupHost(w.host)
            group.Done()
        }(w)
    }

    group.Wait()

}

func main() {
	rand.Seed(time.Now().UnixNano())

	t0 := time.Now()
	lookup()

	fmt.Printf("\n")
	for _, w := range worklist {
		if w.err != nil {
			fmt.Printf("%s: error: %v\n", w.host, w.err)
			continue
		}
		fmt.Printf("%s: %v\n", w.host, w.addrs)
	}
	fmt.Printf("total lookup time: %.3f seconds\n", time.Since(t0).Seconds())
}

var worklist = []*Work{
	{host: "fast.com"},
	{host: "slow.com"},
	{host: "fast.missing.com"},
	{host: "slow.missing.com"},
}

type Work struct {
	host  string
	addrs []string
	err   error
}

func LookupHost(name string) (addrs []string, err error) {
	t0 := time.Now()
	defer func() {
		fmt.Printf("lookup %s: %.3f seconds\n", name, time.Since(t0).Seconds())
	}()
	h := hosts[name]
	if h == nil {
		h = failure
	}
	return h(name)
}

type resolver func(string) ([]string, error)

var hosts = map[string]resolver{
	"fast.com":         delay(10*time.Millisecond, fixedAddrs("10.0.0.1")),
	"slow.com":         delay(2*time.Second, fixedAddrs("10.0.0.4")),
	"fast.missing.com": delay(10*time.Millisecond, failure),
	"slow.missing.com": delay(2*time.Second, failure),
}

func fixedAddrs(addrs ...string) resolver {
	return func(string) ([]string, error) {
		return addrs, nil
	}
}

func delay(d time.Duration, f resolver) resolver {
	return func(name string) ([]string, error) {
		time.Sleep(d/2 + time.Duration(rand.Int63n(int64(d/2))))
		return f(name)
	}
}

func failure(name string) ([]string, error) {
	return nil, fmt.Errorf("unknown host %v", name)
}
35

Concurrency

Aside: can abstract this pattern further (hypothetical): 

var par ParallelDo

for _, w := range worklist {
    w := w // copy iteration variable
    par.Do(func() {
        w.addrs, w.err = net.LookupHost(w.host)            
    })
)

par.Wait()

But it's still useful to be able to construct alternate patterns.

36

Concurrency

Bounded parallelism: 

// +build ignore,OMIT

package main

import (
	"fmt"
	"math/rand"
	"time"
)

func lookup() {

    const max = 2

    done := make(chan bool, len(worklist))
    limit := make(chan bool, max)

    for _, w := range worklist {
        go func(w *Work) {
            limit <- true
            w.addrs, w.err = LookupHost(w.host)
            <-limit
            done <- true
        }(w)
    }

    for i := 0; i < len(worklist); i++ {
        <-done
    }

}

func main() {
	rand.Seed(time.Now().UnixNano())

	t0 := time.Now()
	lookup()

	fmt.Printf("\n")
	for _, w := range worklist {
		if w.err != nil {
			fmt.Printf("%s: error: %v\n", w.host, w.err)
			continue
		}
		fmt.Printf("%s: %v\n", w.host, w.addrs)
	}
	fmt.Printf("total lookup time: %.3f seconds\n", time.Since(t0).Seconds())
}

var worklist = []*Work{
	{host: "fast.com"},
	{host: "slow.com"},
	{host: "fast.missing.com"},
	{host: "slow.missing.com"},
}

type Work struct {
	host  string
	addrs []string
	err   error
}

func LookupHost(name string) (addrs []string, err error) {
	t0 := time.Now()
	defer func() {
		fmt.Printf("lookup %s: %.3f seconds\n", name, time.Since(t0).Seconds())
	}()
	h := hosts[name]
	if h == nil {
		h = failure
	}
	return h(name)
}

type resolver func(string) ([]string, error)

var hosts = map[string]resolver{
	"fast.com":         delay(10*time.Millisecond, fixedAddrs("10.0.0.1")),
	"slow.com":         delay(2*time.Second, fixedAddrs("10.0.0.4")),
	"fast.missing.com": delay(10*time.Millisecond, failure),
	"slow.missing.com": delay(2*time.Second, failure),
}

func fixedAddrs(addrs ...string) resolver {
	return func(string) ([]string, error) {
		return addrs, nil
	}
}

func delay(d time.Duration, f resolver) resolver {
	return func(name string) ([]string, error) {
		time.Sleep(d/2 + time.Duration(rand.Int63n(int64(d/2))))
		return f(name)
	}
}

func failure(name string) ([]string, error) {
	return nil, fmt.Errorf("unknown host %v", name)
}
37

Concurrency

Bounded parallelism, 2: 

// +build ignore,OMIT

package main

import (
	"fmt"
	"math/rand"
	"time"
)

func lookup() {

    const max = 2

    n := 0
    done := make(chan bool, max)

    for _, w := range worklist {
        if n++; n > max {
            <-done
            n--
        }
        go func(w *Work) {
            w.addrs, w.err = LookupHost(w.host)
            done <- true
        }(w)
    }
    for ; n > 0; n-- {
        <-done
    }

}

func main() {
	rand.Seed(time.Now().UnixNano())

	t0 := time.Now()
	lookup()

	fmt.Printf("\n")
	for _, w := range worklist {
		if w.err != nil {
			fmt.Printf("%s: error: %v\n", w.host, w.err)
			continue
		}
		fmt.Printf("%s: %v\n", w.host, w.addrs)
	}
	fmt.Printf("total lookup time: %.3f seconds\n", time.Since(t0).Seconds())
}

var worklist = []*Work{
	{host: "fast.com"},
	{host: "slow.com"},
	{host: "fast.missing.com"},
	{host: "slow.missing.com"},
}

type Work struct {
	host  string
	addrs []string
	err   error
}

func LookupHost(name string) (addrs []string, err error) {
	t0 := time.Now()
	defer func() {
		fmt.Printf("lookup %s: %.3f seconds\n", name, time.Since(t0).Seconds())
	}()
	h := hosts[name]
	if h == nil {
		h = failure
	}
	return h(name)
}

type resolver func(string) ([]string, error)

var hosts = map[string]resolver{
	"fast.com":         delay(10*time.Millisecond, fixedAddrs("10.0.0.1")),
	"slow.com":         delay(2*time.Second, fixedAddrs("10.0.0.4")),
	"fast.missing.com": delay(10*time.Millisecond, failure),
	"slow.missing.com": delay(2*time.Second, failure),
}

func fixedAddrs(addrs ...string) resolver {
	return func(string) ([]string, error) {
		return addrs, nil
	}
}

func delay(d time.Duration, f resolver) resolver {
	return func(name string) ([]string, error) {
		time.Sleep(d/2 + time.Duration(rand.Int63n(int64(d/2))))
		return f(name)
	}
}

func failure(name string) ([]string, error) {
	return nil, fmt.Errorf("unknown host %v", name)
}
38

Concurrency:

Aside: can abstract (still hypothetical): 

par.Limit(10)

for _, w := range work {
    w := w // copy iteration variable
    par.Do(func() {
        w.addrs, w.err = net.LookupHost(w.host)            
    })
)

par.Wait()
39

Concurrency

Example: replicated storage with read and write quorums. 

const (
    F = 2
    N = 5 // >= 2F + 1
    ReadQuorum = F + 1
    WriteQuorum = N - F
)
40

Concurrency

Replicated write, returning after enough writes have succeeded. 

// +build ignore,OMIT

package main

import (
	"fmt"
	"math"
	"math/rand"
	"sync"
	"time"
)

const (
	F           = 2
	N           = 5
	ReadQuorum  = F + 1
	WriteQuorum = N - F
)

var delay = false

type Server struct {
	mu   sync.Mutex
	data map[string]*Data
}

type Data struct {
	Key   string
	Value string
	Time  time.Time
}

func (srv *Server) Delay() {
	if delay == false {
		return
	}
	time.Sleep(time.Duration(math.Abs(rand.NormFloat64()*1e9 + 0.1e9)))
}

func (srv *Server) Write(req *Data) {
	t0 := time.Now()
	defer func() {
		if delay {
			fmt.Printf("write took %.3f seconds\n", time.Since(t0).Seconds())
		}
	}()

	srv.mu.Lock()
	defer srv.mu.Unlock()
	srv.Delay()

	if srv.data == nil {
		srv.data = make(map[string]*Data)
	}
	if d := srv.data[req.Key]; d == nil || d.Time.Before(req.Time) {
		srv.data[req.Key] = req
	}
}

func (srv *Server) Read(key string) *Data {
	t0 := time.Now()
	defer func() {
		fmt.Printf("read took %.3f seconds\n", time.Since(t0).Seconds())
	}()

	srv.mu.Lock()
	defer srv.mu.Unlock()
	srv.Delay()

	return srv.data[key]
}

func better(x, y *Data) *Data {
	if x == nil {
		return y
	}
	if y == nil || y.Time.Before(x.Time) {
		return x
	}
	return y
}

func Write(req *Data) {
	t0 := time.Now()

    done := make(chan bool, len(servers))

    for _, srv := range servers {
        go func(srv *Server) {
            srv.Write(req)
            done <- true
        }(srv)
    }

    for n := 0; n < WriteQuorum; n++ {
        <-done
    }

	if delay {
		fmt.Printf("write committed at %.3f seconds\n", time.Since(t0).Seconds())
	}
	for n := WriteQuorum; n < N; n++ {
		<-done
	}
	if delay {
		fmt.Printf("all replicas written at %.3f seconds\n", time.Since(t0).Seconds())
	}
}

func Read(key string) {
	t0 := time.Now()
	replies := make(chan *Data, len(servers))

	for _, srv := range servers {
		go func(srv *Server) {
			replies <- srv.Read(key)
		}(srv)
	}

	var d *Data
	for n := 0; n < ReadQuorum; n++ {
		d = better(d, <-replies)
	}

	if delay {
		fmt.Printf("read committed at %.3f seconds\n", time.Since(t0).Seconds())
	}

	for n := ReadQuorum; n < N; n++ {
		<-replies
	}
	if delay {
		fmt.Printf("all replicas read at %.3f seconds\n", time.Since(t0).Seconds())
	}
}

var servers []*Server

func main() {
	servers = make([]*Server, N)
	for i := range servers {
		servers[i] = new(Server)
	}

	rand.Seed(time.Now().UnixNano())

	delay = false
	Write(&Data{"hello", "there", time.Now()})
	time.Sleep(1 * time.Millisecond)

	delay = true
	Write(&Data{"hello", "world", time.Now()})

	//	Read("hello")
}
41

Concurrency

Replicated read, returning after enough reads have been gathered. 

// +build ignore,OMIT

package main

import (
	"fmt"
	"math"
	"math/rand"
	"sync"
	"time"
)

const (
	F           = 2
	N           = 5
	ReadQuorum  = F + 1
	WriteQuorum = N - F
)

var delay = false

type Server struct {
	mu   sync.Mutex
	data map[string]*Data
}

type Data struct {
	Key   string
	Value string
	Time  time.Time
}

func (srv *Server) Delay() {
	if delay == false {
		return
	}
	time.Sleep(time.Duration(math.Abs(rand.NormFloat64()*1e9 + 0.1e9)))
}

func (srv *Server) Write(req *Data) {
	t0 := time.Now()
	defer func() {
		if delay {
			fmt.Printf("write took %.3f seconds\n", time.Since(t0).Seconds())
		}
	}()

	srv.mu.Lock()
	defer srv.mu.Unlock()
	srv.Delay()

	if srv.data == nil {
		srv.data = make(map[string]*Data)
	}
	if d := srv.data[req.Key]; d == nil || d.Time.Before(req.Time) {
		srv.data[req.Key] = req
	}
}

func (srv *Server) Read(key string) *Data {
	t0 := time.Now()
	defer func() {
		fmt.Printf("read took %.3f seconds\n", time.Since(t0).Seconds())
	}()

	srv.mu.Lock()
	defer srv.mu.Unlock()
	srv.Delay()

	return srv.data[key]
}

func better(x, y *Data) *Data {
	if x == nil {
		return y
	}
	if y == nil || y.Time.Before(x.Time) {
		return x
	}
	return y
}

func Write(req *Data) {
	t0 := time.Now()
	done := make(chan bool, len(servers))

	for _, srv := range servers {
		go func(srv *Server) {
			srv.Write(req)
			done <- true
		}(srv)
	}

	for n := 0; n < WriteQuorum; n++ {
		<-done
	}

	if delay {
		fmt.Printf("write committed at %.3f seconds\n", time.Since(t0).Seconds())
	}
	for n := WriteQuorum; n < N; n++ {
		<-done
	}
	if delay {
		fmt.Printf("all replicas written at %.3f seconds\n", time.Since(t0).Seconds())
	}
}

func Read(key string) {
	t0 := time.Now()

    replies := make(chan *Data, len(servers))

    for _, srv := range servers {
        go func(srv *Server) {
            replies <- srv.Read(key)
        }(srv)
    }

    var d *Data
    for n := 0; n < ReadQuorum; n++ {
        d = better(d, <-replies)
    }

	if delay {
		fmt.Printf("read committed at %.3f seconds\n", time.Since(t0).Seconds())
	}

	for n := ReadQuorum; n < N; n++ {
		<-replies
	}
	if delay {
		fmt.Printf("all replicas read at %.3f seconds\n", time.Since(t0).Seconds())
	}
}

var servers []*Server

func main() {
	servers = make([]*Server, N)
	for i := range servers {
		servers[i] = new(Server)
	}

	rand.Seed(time.Now().UnixNano())

	delay = false
	Write(&Data{"hello", "there", time.Now()})
	time.Sleep(1 * time.Millisecond)

	Write(&Data{"hello", "world", time.Now()})

	delay = true
	Read("hello")
}
42

Concurrency

Select allows choosing between multiple channel operations.

Example, chat program: 

for {
    select {
    case event := <-ui:
        // process user interface event
    case msg := <-server:
        // process server message
    case t := <-tick:
        // time has elapsed
    }
}
43

Concurrency Lessons

Most important:

44

Production Use

vitess/vtocc, MySQL query balancer

groupcache

45

More information and related talks

https://go.dev/

rsc@golang.org

Videos:

Questions?

46

Thank you

Russ Cox

Google

Use the left and right arrow keys or click the left and right edges of the page to navigate between slides.
(Press 'H' or navigate to hide this message.)