Add feature for graceful shutdown of HTTP server

If the watchdog is sent SIGTERM from an external process then it
should stop accepting new connections and attempt to finish the
work in progress. This change makes use of the new ability in Go
1.9 and onwards to cancel a HTTP server gracefully.

The write_timeout duration is used as a grace period to allow all
in-flight requests to complete. The pattern is taken directly from
the offical example in the Golang documentation. [1]

Further tuning and testing may be needed for Windows containers which
have a different set of signals for closing work. This change aims
to cover the majority use-case for Linux containers.

The HTTP health-check is also invalidated by creating an and
expression with the existing lock file.

Tested with Kubernetes by deploying a custom watchdog and the
fprocess of `env`. Log message was observed when scaling down and
connections stopped being accepted on terminating replica.

Also corrects some typos from previous PR.

[1] https://golang.org/pkg/net/http/#Server.Shutdown

Signed-off-by: Alex Ellis (VMware) <alexellis2@gmail.com>

watchdog/main.go

@@ -5,20 +5,25 @@ package main
 
 import (
 	"bytes"
+	"context"
 	"fmt"
 	"io/ioutil"
 	"log"
 	"net/http"
 	"os"
 	"os/exec"
+	"os/signal"
 	"path/filepath"
 	"strings"
 	"sync"
+	"syscall"
 	"time"
 
 	"github.com/openfaas/faas/watchdog/types"
 )
 
+var acceptingConnections bool
+
 // buildFunctionInput for a GET method this is an empty byte array.
 func buildFunctionInput(config *WatchdogConfig, r *http.Request) ([]byte, error) {
 	var res []byte
@@ -260,25 +265,20 @@ func lockFilePresent() bool {
 	return true
 }
 
-func createLockFile() error {
+func createLockFile() (string, error) {
 	path := filepath.Join(os.TempDir(), ".lock")
 	log.Printf("Writing lock-file to: %s\n", path)
 	writeErr := ioutil.WriteFile(path, []byte{}, 0660)
-	return writeErr
+	acceptingConnections = true
-}
 
-func removeLockFile() error {
+	return path, writeErr
-	path := filepath.Join(os.TempDir(), ".lock")
-	log.Printf("Removing lock-file : %s\n", path)
-	removeErr := os.Remove(path)
-	return removeErr
 }
 
 func makeHealthHandler() func(http.ResponseWriter, *http.Request) {
 	return func(w http.ResponseWriter, r *http.Request) {
 		switch r.Method {
 		case http.MethodGet:
-			if lockFilePresent() == false {
+			if acceptingConnections == false || lockFilePresent() == false {
 				w.WriteHeader(http.StatusInternalServerError)
 				return
 			}
@@ -288,9 +288,7 @@ func makeHealthHandler() func(http.ResponseWriter, *http.Request) {
 			break
 		default:
 			w.WriteHeader(http.StatusMethodNotAllowed)
-
 		}
-
 	}
 }
 
@@ -312,6 +310,8 @@ func makeRequestHandler(config *WatchdogConfig) func(http.ResponseWriter, *http.
 }
 
 func main() {
+	acceptingConnections = false
+
 	osEnv := types.OsEnv{}
 	readConfig := ReadConfig{}
 	config := readConfig.Read(osEnv)
@@ -335,15 +335,46 @@ func main() {
 	http.HandleFunc("/", makeRequestHandler(&config))
 
 	if config.suppressLock == false {
-		path := filepath.Join(os.TempDir(), ".lock")
+		path, writeErr := createLockFile()
-		log.Printf("Writing lock-file to: %s\n", path)
+
-		writeErr := ioutil.WriteFile(path, []byte{}, 0660)
 		if writeErr != nil {
 			log.Panicf("Cannot write %s. To disable lock-file set env suppress_lock=true.\n Error: %s.\n", path, writeErr.Error())
 		}
 	} else {
 		log.Println("Warning: \"suppress_lock\" is enabled. No automated health-checks will be in place for your function.")
+		acceptingConnections = true
 	}
 
-	log.Fatal(s.ListenAndServe())
+	listenUntilShutdown(config.writeTimeout, s)
+}
+
+func listenUntilShutdown(shutdownTimeout time.Duration, s *http.Server) {
+
+	idleConnsClosed := make(chan struct{})
+	go func() {
+		sig := make(chan os.Signal, 1)
+		signal.Notify(sig, syscall.SIGTERM)
+
+		<-sig
+
+		log.Printf("SIGTERM received.. shutting down server")
+
+		acceptingConnections = false
+
+		if err := s.Shutdown(context.Background()); err != nil {
+			// Error from closing listeners, or context timeout:
+			log.Printf("Error in Shutdown: %v", err)
+		}
+
+		<-time.Tick(shutdownTimeout)
+
+		close(idleConnsClosed)
+	}()
+
+	if err := s.ListenAndServe(); err != http.ErrServerClosed {
+		log.Printf("Error ListenAndServe: %v", err)
+		close(idleConnsClosed)
+	}
+
+	<-idleConnsClosed
 }

watchdog/requesthandler_test.go

@@ -11,6 +11,7 @@ import (
 	"net/http"
 	"net/http/httptest"
 	"os"
+	"path/filepath"
 	"strings"
 	"testing"
 	"time"
@@ -361,11 +362,13 @@ func TestHandler_StatusOKForGETAndNoBody(t *testing.T) {
 	}
 }
 
-func TestHealthHandler_SatusOK_LockFilePresent(t *testing.T) {
+func TestHealthHandler_StatusOK_LockFilePresent(t *testing.T) {
 	rr := httptest.NewRecorder()
 
-	if lockFilePresent() == false {
+	present := lockFilePresent()
-		if err := createLockFile(); err != nil {
+
+	if present == false {
+		if _, err := createLockFile(); err != nil {
 			t.Fatal(err)
 		}
 	}
@@ -402,7 +405,7 @@ func TestHealthHandler_StatusInternalServerError_LockFileNotPresent(t *testing.T
 
 	required := http.StatusInternalServerError
 	if status := rr.Code; status != required {
-		t.Errorf("handler retruned wrong status code: got %v, but wanted %v", status, required)
+		t.Errorf("handler returned wrong status code - got: %v, want: %v", status, required)
 	}
 }
 
@@ -426,3 +429,10 @@ func TestHealthHandler_SatusMethoNotAllowed_ForWriteableVerbs(t *testing.T) {
 		}
 	}
 }
+
+func removeLockFile() error {
+	path := filepath.Join(os.TempDir(), ".lock")
+	log.Printf("Removing lock-file : %s\n", path)
+	removeErr := os.Remove(path)
+	return removeErr
+}