This question already has answers here:
fatal error: all goroutines are asleep - deadlock! when channel is not buffered
(1 answer)
Why is the order of channels receiving causing/resolving a deadlock in Golang?
(3 answers)
Closed 4 months ago.
I wrote the following code trying to implement future for a group of async jobs. and it works. but I don't understand why I have to wrap wg.Wait() in a goroutine w/o it main will block.
package main
import (
"fmt"
"net/http"
"sync"
)
var urls = []string{
"https://www.easyjet.com/",
"https://www.skyscanner.de/",
"https://www.ryanair.com",
"https://wizzair.com/",
"https://www.swiss.com/",
}
func checker(urls []string) func() chan urlStatus {
wg := &sync.WaitGroup{}
c := make(chan urlStatus)
for _, url := range urls {
wg.Add(1)
go checkUrl(url, c, wg)
}
return func() chan urlStatus {
go func() { // <--- why this goroutine is necessary. change it to sync will block
wg.Wait()
close(c)
}()
return c
}
}
func main() {
done := make(chan bool)
future := checker(urls)
go consumer(future(), done)
<-done
}
func consumer(c <-chan urlStatus, done chan bool) {
for r := range c {
if r.status {
fmt.Println(r.url, "is up.")
} else {
fmt.Println(r.url, "is down!!")
}
}
fmt.Println("end of consumer")
done <- true
}
//checks and prints a message if a website is up or down
func checkUrl(url string, c chan<- urlStatus, wg *sync.WaitGroup) {
defer wg.Done()
_, err := http.Get(url)
if err != nil {
// The website is down
c <- urlStatus{url, false}
} else {
// The website is up
c <- urlStatus{url, true}
}
}
type urlStatus struct {
url string
status bool
}
I think it is because of this line
go consumer(future(), done)
When future() is called, the main go routine is blocked by wg.Wait()
Related
I am having issue while using waitgroup with the buffered channel. The problem is waitgroup closes before channel is read completely, which make my channel is half read and break in between.
func main() {
var wg sync.WaitGroup
var err error
start := time.Now()
students := make([]studentDetails, 0)
studentCh := make(chan studentDetail, 10000)
errorCh := make(chan error, 1)
wg.Add(1)
go s.getDetailStudents(rCtx, studentCh , errorCh, &wg, s.Link, false)
go func(ch chan studentDetail, e chan error) {
LOOP:
for {
select {
case p, ok := <-ch:
if ok {
L.Printf("Links %s: [%s]\n", p.title, p.link)
students = append(students, p)
} else {
L.Print("Closed channel")
break LOOP
}
case err = <-e:
if err != nil {
break
}
}
}
}(studentCh, errorCh)
wg.Wait()
close(studentCh)
close(errorCh)
L.Warnln("closed: all wait-groups completed!")
L.Warnf("total items fetched: %d", len(students))
elapsed := time.Since(start)
L.Warnf("operation took %s", elapsed)
}
The problem is this function is recursive. I mean some http call to fetch students and then make more calls depending on condition.
func (s Student) getDetailStudents(rCtx context.Context, content chan<- studentDetail, errorCh chan<- error, wg *sync.WaitGroup, url string, subSection bool) {
util.MustNotNil(rCtx)
L := logger.GetLogger(rCtx)
defer func() {
L.Println("Closing all waitgroup!")
wg.Done()
}()
wc := getWC()
httpClient := wc.Registry.MustHTTPClient()
res, err := httpClient.Get(url)
if err != nil {
L.Fatal(err)
}
defer res.Body.Close()
if res.StatusCode != 200 {
L.Errorf("status code error: %d %s", res.StatusCode, res.Status)
errorCh <- errors.New("service_status_code")
return
}
// parse response and return error if found some through errorCh as done above.
// decide page subSection based on response if it is more.
if !subSection {
wg.Add(1)
go s.getDetailStudents(rCtx, content, errorCh, wg, link, true)
// L.Warnf("total pages found %d", pageSub.Length()+1)
}
// Find students from response list and parse each Student
students := s.parseStudentItemList(rCtx, item)
for _, student := range students {
content <- student
}
L.Warnf("Calling HTTP Service for %q with total %d record", url, elementsSub.Length())
}
Variables are changed to avoid original code base.
The problem is students are read randomly as soon as Waitgroup complete. I am expecting to hold the execution until all students are read, In case of error it should break as soon error encounter.
You need to know when the receiving goroutine completes. The WaitGroup does that for the generating goroutine. So, you can use two waitgroups:
wg.Add(1)
go s.getDetailStudents(rCtx, studentCh , errorCh, &wg, s.Link, false)
wgReader.Add(1)
go func(ch chan studentDetail, e chan error) {
defer wgReader.Done()
...
}
wg.Wait()
close(studentCh)
close(errorCh)
wgReader.Wait() // Wait for the readers to complete
Since you are using buffered channels you can retrieve the remaining values after closing the channel. You will also need a mechanism to prevent your main function from exiting too early while the reader is still doing work ,as #Burak Serdar has advised.
I restructured the code to give a working example but it should get the point across.
package main
import (
"context"
"log"
"sync"
"time"
)
type studentDetails struct {
title string
link string
}
func main() {
var wg sync.WaitGroup
var err error
students := make([]studentDetails, 0)
studentCh := make(chan studentDetails, 10000)
errorCh := make(chan error, 1)
start := time.Now()
wg.Add(1)
go getDetailStudents(context.TODO(), studentCh, errorCh, &wg, "http://example.com", false)
LOOP:
for {
select {
case p, ok := <-studentCh:
if ok {
log.Printf("Links %s: [%s]\n", p.title, p.link)
students = append(students, p)
} else {
log.Println("Draining student channel")
for p := range studentCh {
log.Printf("Links %s: [%s]\n", p.title, p.link)
students = append(students, p)
}
break LOOP
}
case err = <-errorCh:
if err != nil {
break LOOP
}
case <-wrapWait(&wg):
close(studentCh)
}
}
close(errorCh)
elapsed := time.Since(start)
log.Printf("operation took %s", elapsed)
}
func getDetailStudents(rCtx context.Context, content chan<- studentDetails, errorCh chan<- error, wg *sync.WaitGroup, url string, subSection bool) {
defer func() {
log.Println("Closing")
wg.Done()
}()
if !subSection {
wg.Add(1)
go getDetailStudents(rCtx, content, errorCh, wg, url, true)
// L.Warnf("total pages found %d", pageSub.Length()+1)
}
content <- studentDetails{
title: "title",
link: "link",
}
}
// helper function to allow using WaitGroup in a select
func wrapWait(wg *sync.WaitGroup) <-chan struct{} {
out := make(chan struct{})
go func() {
wg.Wait()
out <- struct{}{}
}()
return out
}
wg.Add(1)
go func(){
defer wg.Done()
// I do not think that you need a recursive function.
// this function overcomplicated.
s.getDetailStudents(rCtx, studentCh , errorCh, &wg, s.Link, false)
}(...)
wg.Add(1)
go func(ch chan studentDetail, e chan error) {
defer wg.Done()
...
}(...)
wg.Wait()
close(studentCh)
close(errorCh)
This should solve the problem. s.getDetailStudents function must be simplified. Making it recursive does not have any benefit.
I try to send an error in the channel on recovery
Why this error is not sent to the channel?
package main
import (
"fmt"
"sync"
"errors"
)
func main() {
var wg sync.WaitGroup
wg.Add(1)
batchErrChan := make(chan error)
go func(errchan chan error) {
defer func() {
if r := recover(); r != nil {
errchan <- errors.New("recover err")
}
close(batchErrChan)
wg.Done()
}()
panic("ddd")
}(batchErrChan)
go func() {
for _ = range batchErrChan {
fmt.Println("err in range")
}
}()
wg.Wait()
}
https://play.golang.org/p/0ytunuYDWZU
I expect "err in range" to be printed, but it is not. Why?
Your program ends before the goroutine gets a chance to print the message. Try waiting to it:
...
done:=make(chan struct{})
go func() {
for _ = range batchErrChan {
fmt.Println("err in range")
}
close(done)
}()
wg.Wait()
<-done
}
I am playing around with Golang and I created this little app to make several concurrent api calls using goroutines.
While the app works, after the calls complete, the app gets stuck, which makes sense because it cannot exit the range c loop because the channel is not closed.
I am not sure where to better close the channel in this pattern.
package main
import "fmt"
import "net/http"
func main() {
links := []string{
"https://github.com/fabpot",
"https://github.com/andrew",
"https://github.com/taylorotwell",
"https://github.com/egoist",
"https://github.com/HugoGiraudel",
}
checkUrls(links)
}
func checkUrls(urls []string) {
c := make(chan string)
for _, link := range urls {
go checkUrl(link, c)
}
for msg := range c {
fmt.Println(msg)
}
close(c) //this won't get hit
}
func checkUrl(url string, c chan string) {
_, err := http.Get(url)
if err != nil {
c <- "We could not reach:" + url
} else {
c <- "Success reaching the website:" + url
}
}
You close a channel when there are no more values to send, so in this case it's when all checkUrl goroutines have completed.
var wg sync.WaitGroup
func checkUrls(urls []string) {
c := make(chan string)
for _, link := range urls {
wg.Add(1)
go checkUrl(link, c)
}
go func() {
wg.Wait()
close(c)
}()
for msg := range c {
fmt.Println(msg)
}
}
func checkUrl(url string, c chan string) {
defer wg.Done()
_, err := http.Get(url)
if err != nil {
c <- "We could not reach:" + url
} else {
c <- "Success reaching the website:" + url
}
}
(Note that the error from http.Get is only going to reflect connection and protocol errors. It is not going to contain http server errors if you're expecting those too, which you must be seeing how you're checking for paths and not just hosts.)
When writing programs in Go using channels and goroutines always think about who (which function) owns a channel. I prefer the practice of letting the function who owns a channel close it. If i were to write this i would do as shown below.
Note: A better way to handle situations like this is the Fan-out, fan-in concurrency pattern. refer(https://blog.golang.org/pipelines)Go Concurrency Patterns
package main
import "fmt"
import "net/http"
import "sync"
func main() {
links := []string{
"https://github.com/fabpot",
"https://github.com/andrew",
"https://github.com/taylorotwell",
"https://github.com/egoist",
"https://github.com/HugoGiraudel",
}
processURLS(links)
fmt.Println("End of Main")
}
func processURLS(links []string) {
resultsChan := checkUrls(links)
for msg := range resultsChan {
fmt.Println(msg)
}
}
func checkUrls(urls []string) chan string {
outChan := make(chan string)
go func(urls []string) {
defer close(outChan)
var wg sync.WaitGroup
for _, url := range urls {
wg.Add(1)
go checkUrl(&wg, url, outChan)
}
wg.Wait()
}(urls)
return outChan
}
func checkUrl(wg *sync.WaitGroup, url string, c chan string) {
defer wg.Done()
_, err := http.Get(url)
if err != nil {
c <- "We could not reach:" + url
} else {
c <- "Success reaching the website:" + url
}
}
I have the following golang program;
package main
import (
"fmt"
"net/http"
"time"
)
var urls = []string{
"http://www.google.com/",
"http://golang.org/",
"http://yahoo.com/",
}
type HttpResponse struct {
url string
response *http.Response
err error
status string
}
func asyncHttpGets(url string, ch chan *HttpResponse) {
client := http.Client{}
if url == "http://www.google.com/" {
time.Sleep(500 * time.Millisecond) //google is down
}
fmt.Printf("Fetching %s \n", url)
resp, err := client.Get(url)
u := &HttpResponse{url, resp, err, "fetched"}
ch <- u
fmt.Println("sent to chan")
}
func main() {
fmt.Println("start")
ch := make(chan *HttpResponse, len(urls))
for _, url := range urls {
go asyncHttpGets(url, ch)
}
for i := range ch {
fmt.Println(i)
}
fmt.Println("Im done")
}
Run it on Playground
However when I run it; it hangs (ie the last part that ought to print Im done doesnt run.)
Here's the terminal output;;
$ go run get.go
start
Fetching http://yahoo.com/
Fetching http://golang.org/
Fetching http://www.google.com/
sent to chan
&{http://www.google.com/ 0xc820144120 fetched}
sent to chan
&{http://golang.org/ 0xc82008b710 fetched}
sent to chan
&{http://yahoo.com/ 0xc82008b7a0 fetched}
The problem is that ranging over a channel in a for loop will continue forever unless the channel is closed. If you want to read precisely len(urls) values from the channel, you should loop that many times:
for i := 0; i < len(urls); i++ {
fmt.Println(<-ch)
}
Another good dirty devious trick would be to use sync.WaitGroup and increment it per goroutine and then monitor it with a Wait and after its done it will close your channel allowing the next blocks of code to run, the reason I am offering you this approach is because it gets away from using a static number in a loop like len(urls) so that you can have a dynamic slice that might change and what not.
The reason Wait and close are in their own goroutine is so that your code can reach the for loop to range over your channel
package main
import (
"fmt"
"net/http"
"time"
"sync"
)
var urls = []string{
"http://www.google.com/",
"http://golang.org/",
"http://yahoo.com/",
}
type HttpResponse struct {
url string
response *http.Response
err error
status string
}
func asyncHttpGets(url string, ch chan *HttpResponse, wg *sync.WaitGroup) {
client := http.Client{}
if url == "http://www.google.com/" {
time.Sleep(500 * time.Millisecond) //google is down
}
fmt.Printf("Fetching %s \n", url)
resp, err := client.Get(url)
u := &HttpResponse{url, resp, err, "fetched"}
ch <- u
fmt.Println("sent to chan")
wg.Done()
}
func main() {
fmt.Println("start")
ch := make(chan *HttpResponse, len(urls))
var wg sync.WaitGroup
for _, url := range urls {
wg.Add(1)
go asyncHttpGets(url, ch, &wg)
}
go func() {
wg.Wait()
close(ch)
}()
for i := range ch {
fmt.Println(i)
}
fmt.Println("Im done")
}
I got a problem using sync.WaitGroup and select together. If you take a look at following http request pool you will notice that if an error occurs it will never be reported as wg.Done() will block and there is no read from the channel anymore.
package pool
import (
"fmt"
"log"
"net/http"
"sync"
)
var (
MaxPoolQueue = 100
MaxPoolWorker = 10
)
type Pool struct {
wg *sync.WaitGroup
queue chan *http.Request
errors chan error
}
func NewPool() *Pool {
return &Pool{
wg: &sync.WaitGroup{},
queue: make(chan *http.Request, MaxPoolQueue),
errors: make(chan error),
}
}
func (p *Pool) Add(r *http.Request) {
p.wg.Add(1)
p.queue <- r
}
func (p *Pool) Run() error {
for i := 0; i < MaxPoolWorker; i++ {
go p.doWork()
}
select {
case err := <-p.errors:
return err
default:
p.wg.Wait()
}
return nil
}
func (p *Pool) doWork() {
for r := range p.queue {
fmt.Printf("Request to %s\n", r.Host)
p.wg.Done()
_, err := http.DefaultClient.Do(r)
if err != nil {
log.Fatal(err)
p.errors <- err
} else {
fmt.Printf("no error\n")
}
}
}
Source can be found here
How can I still use WaitGroup but also get errors from go routines?
Just got the answer my self as I wrote the question and as I think it is an interesting case I would like to share it with you.
The trick to use sync.WaitGroup and chan together is that we wrap:
select {
case err := <-p.errors:
return err
default:
p.wg.Done()
}
Together in a for loop:
for {
select {
case err := <-p.errors:
return err
default:
p.wg.Done()
}
}
In this case select will always check for errors and wait if nothing happens :)
It looks a bit like the fail-fast mechanism enabled by the Tomb library (Tomb V2 GoDoc):
The tomb package handles clean goroutine tracking and termination.
If any of the tracked goroutines returns a non-nil error, or the Kill or Killf method is called by any goroutine in the system (tracked or not), the tomb Err is set, Alive is set to false, and the Dying channel is closed to flag that all tracked goroutines are supposed to willingly terminate as soon as possible.
Once all tracked goroutines terminate, the Dead channel is closed, and Wait unblocks and returns the first non-nil error presented to the tomb via a result or an explicit Kill or Killf method call, or nil if there were no errors.
You can see an example in this playground:
(extract)
// start runs all the given functions concurrently
// until either they all complete or one returns an
// error, in which case it returns that error.
//
// The functions are passed a channel which will be closed
// when the function should stop.
func start(funcs []func(stop <-chan struct{}) error) error {
var tomb tomb.Tomb
var wg sync.WaitGroup
allDone := make(chan struct{})
// Start all the functions.
for _, f := range funcs {
f := f
wg.Add(1)
go func() {
defer wg.Done()
if err := f(tomb.Dying()); err != nil {
tomb.Kill(err)
}
}()
}
// Start a goroutine to wait for them all to finish.
go func() {
wg.Wait()
close(allDone)
}()
// Wait for them all to finish, or one to fail
select {
case <-allDone:
case <-tomb.Dying():
}
tomb.Done()
return tomb.Err()
}
A simpler implementation would be like below. (Check in play.golang: https://play.golang.org/p/TYxxsDRt5Wu)
package main
import "fmt"
import "sync"
import "time"
type Error struct {
message string
}
func (e Error) Error() string {
return e.message
}
func main() {
var wg sync.WaitGroup
waitGroupLength := 8
errChannel := make(chan error, 1)
// Setup waitgroup to match the number of go routines we'll launch off
wg.Add(waitGroupLength)
finished := make(chan bool, 1) // this along with wg.Wait() are why the error handling works and doesn't deadlock
for i := 0; i < waitGroupLength; i++ {
go func(i int) {
fmt.Printf("Go routine %d executed\n", i+1)
time.Sleep(time.Duration(waitGroupLength - i))
time.Sleep(0) // only here so the time import is needed
if i%4 == 1 {
errChannel <- Error{fmt.Sprintf("Errored on routine %d", i+1)}
}
// Mark the wait group as Done so it does not hang
wg.Done()
}(i)
}
go func() {
wg.Wait()
close(finished)
}()
L:
for {
select {
case <-finished:
break L // this will break from loop
case err := <-errChannel:
if err != nil {
fmt.Println("error ", err)
// handle your error
}
}
}
fmt.Println("Executed all go routines")
}