I have written a basic web server in Go, but it can't seem to handle any requests properly and instead all I get is "Recv failure: Connection reset by peer".
I'm happy that the server does indeed start and stop correctly, and I can see it listening on port 8080 that I've configured. I've also ruled out issues with parsing the YAML config file - it's definitely getting getting parsed in to http.Server{}.
I'm not really sure what else to check, and struggling to find anything that's pointing me in the right direction.
I'll also apologise in advance, as I know I'm pasting a large amount of code below, but I really don't know what and where the error is coming from.
Given that the server is running, when I hit the "/" endpoint/route, I'd expect to get "Hello from Go!" returned back.
directory structure
❯ tree . | grep -iE 'cmd|server.go|go.mod|go.sum|server.yaml'
├── cmd
│ └── server.go
├── go.mod
├── go.sum
└── server.yaml
server.go
package main
import (
"context"
"errors"
"fmt"
"log"
"net/http"
"os"
"os/signal"
"syscall"
"time"
"gopkg.in/yaml.v3"
)
type Config struct {
Port string `yaml:"Port"`
ReadTimeout int `yaml:"ReadTimeout"`
WriteTimeout int `yaml:"WriteTimeout"`
IdleTimeout int `yaml:"IdleTimeout"`
ShutdownTimeout int `yaml:"ShutdownTimeout"`
ErrorLog string `yaml:"ErrorLog"`
}
func main() {
if len(os.Args) != 2 {
log.Fatal("Missing arguments")
}
configFile, err := os.Open(os.Args[1])
if err != nil {
log.Fatal(err)
}
defer configFile.Close()
// TODO: Implement a custom ServeMux routes
serverConfig := Config{}
yamlDecoder := yaml.NewDecoder(configFile)
err = yamlDecoder.Decode(&serverConfig)
if err != nil {
log.Fatal(err)
}
errorLogFile, err := os.OpenFile(serverConfig.ErrorLog, os.O_CREATE|os.O_WRONLY|os.O_APPEND, 0666)
if err != nil {
log.Fatal(err)
}
defer errorLogFile.Close()
errorLog := log.New(errorLogFile, "ERROR : ", log.LstdFlags|log.Lshortfile)
server := &http.Server{
Addr: fmt.Sprintf(":%s", serverConfig.Port),
ReadTimeout: time.Duration(serverConfig.ReadTimeout),
WriteTimeout: time.Duration(serverConfig.WriteTimeout),
IdleTimeout: time.Duration(serverConfig.IdleTimeout),
ErrorLog: errorLog,
}
http.HandleFunc("/", func(w http.ResponseWriter, r *http.Request) {
fmt.Fprintln(w, "Hello from Go!")
})
log.Println("Starting the server...")
go func() {
err := server.ListenAndServe()
if !errors.Is(err, http.ErrServerClosed) {
log.Fatal(err)
}
log.Println("Stopped serving new connections")
}()
log.Println("Server started")
sigChan := make(chan os.Signal, 1)
signal.Notify(sigChan, syscall.SIGINT, syscall.SIGTERM)
<-sigChan
ctx, shutdown := context.WithTimeout(context.Background(), time.Duration(serverConfig.ShutdownTimeout)*time.Second)
defer shutdown()
err = server.Shutdown(ctx)
if err != nil {
log.Fatal(err)
}
log.Println("Server gracefully stopped")
}
error message
❯ curl -v localhost:8080/
* Trying ::1...
* TCP_NODELAY set
* Connected to localhost (::1) port 8080 (#0)
> GET / HTTP/1.1
> Host: localhost:8080
> User-Agent: curl/7.64.1
> Accept: */*
>
* Recv failure: Connection reset by peer
* Closing connection 0
curl: (56) Recv failure: Connection reset by peer
CodePudding user response:
As mentioned in the comments, the issue is related to the timeout values being so short - 5 nanoseconds each. This is because time.Duration is represented as the elapsed time between two instants as an int64 nanosecond count. So I needed to convert that in to seconds, to get what I was expecting.
From the docs:
"A Duration represents the elapsed time between two instants as an int64 nanosecond count. The representation limits the largest representable duration to approximately 290 years."
Solution:
server := &http.Server{
Addr: fmt.Sprintf(":%s", serverConfig.Port),
ReadTimeout: time.Duration(serverConfig.ReadTimeout) * time.Second,
WriteTimeout: time.Duration(serverConfig.WriteTimeout) * time.Second,
IdleTimeout: time.Duration(serverConfig.IdleTimeout) * time.Second,
ErrorLog: errorLog,
}
