Add a file lock to the data directory on startup to prevent multiple agents. (#18483) (#18529)

* Add a file lock to the data directory on startup to prevent multiple agents.

* Add export comments to AppLocker.

* Fix periodic to not block startup.

(cherry picked from commit e1a4741)

Author: blakerouse

x-pack/elastic-agent/pkg/agent/application/locker.go

@@ -0,0 +1,50 @@
+// Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+// or more contributor license agreements. Licensed under the Elastic License;
+// you may not use this file except in compliance with the Elastic License.
+
+package application
+
+import (
+	"fmt"
+	"os"
+	"path/filepath"
+
+	"github.com/gofrs/flock"
+)
+
+const lockFileName = "agent.lock"
+
+// ErrAppAlreadyRunning error returned when another elastic-agent is already holding the lock.
+var ErrAppAlreadyRunning = fmt.Errorf("another elastic-agent is already running")
+
+// AppLocker locks the agent.lock file inside the provided directory.
+type AppLocker struct {
+	lock *flock.Flock
+}
+
+// NewAppLocker creates an AppLocker that locks the agent.lock file inside the provided directory.
+func NewAppLocker(dir string) *AppLocker {
+	if _, err := os.Stat(dir); os.IsNotExist(err) {
+		_ = os.Mkdir(dir, 0755)
+	}
+	return &AppLocker{
+		lock: flock.New(filepath.Join(dir, lockFileName)),
+	}
+}
+
+// TryLock tries to grab the lock file and returns error if it cannot.
+func (a *AppLocker) TryLock() error {
+	locked, err := a.lock.TryLock()
+	if err != nil {
+		return err
+	}
+	if !locked {
+		return ErrAppAlreadyRunning
+	}
+	return nil
+}
+
+// Unlock releases the lock file.
+func (a *AppLocker) Unlock() error {
+	return a.lock.Unlock()
+}

x-pack/elastic-agent/pkg/agent/application/locker_test.go

@@ -0,0 +1,29 @@
+// Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+// or more contributor license agreements. Licensed under the Elastic License;
+// you may not use this file except in compliance with the Elastic License.
+
+package application
+
+import (
+	"io/ioutil"
+	"os"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+)
+
+func TestAppLocker(t *testing.T) {
+	tmp, _ := ioutil.TempDir("", "locker")
+	defer os.RemoveAll(tmp)
+
+	locker1 := NewAppLocker(tmp)
+	locker2 := NewAppLocker(tmp)
+
+	require.NoError(t, locker1.TryLock())
+	assert.Error(t, locker2.TryLock())
+	require.NoError(t, locker1.Unlock())
+	require.NoError(t, locker2.TryLock())
+	assert.Error(t, locker1.TryLock())
+	require.NoError(t, locker2.Unlock())
+}

x-pack/elastic-agent/pkg/agent/application/periodic.go

@@ -23,21 +23,24 @@ type periodic struct {
 }
 
 func (p *periodic) Start() error {
-	if err := p.work(); err != nil {
-		p.log.Debugf("Failed to read configuration, error: %s", err)
-	}
-
-	for {
-		select {
-		case <-p.done:
-			break
-		case <-time.After(p.period):
-		}
-
+	go func() {
 		if err := p.work(); err != nil {
 			p.log.Debugf("Failed to read configuration, error: %s", err)
 		}
-	}
+
+		for {
+			select {
+			case <-p.done:
+				break
+			case <-time.After(p.period):
+			}
+
+			if err := p.work(); err != nil {
+				p.log.Debugf("Failed to read configuration, error: %s", err)
+			}
+		}
+	}()
+	return nil
 }
 
 func (p *periodic) work() error {

x-pack/elastic-agent/pkg/agent/cmd/run.go

@@ -13,6 +13,7 @@ import (
 	"github.com/spf13/cobra"
 
 	"github.com/elastic/beats/v7/x-pack/elastic-agent/pkg/agent/application"
+	"github.com/elastic/beats/v7/x-pack/elastic-agent/pkg/agent/application/paths"
 	"github.com/elastic/beats/v7/x-pack/elastic-agent/pkg/agent/errors"
 	"github.com/elastic/beats/v7/x-pack/elastic-agent/pkg/cli"
 	"github.com/elastic/beats/v7/x-pack/elastic-agent/pkg/config"
@@ -47,6 +48,12 @@ func run(flags *globalFlags, streams *cli.IOStreams) error {
 		return err
 	}
 
+	locker := application.NewAppLocker(paths.Data())
+	if err := locker.TryLock(); err != nil {
+		return err
+	}
+	defer locker.Unlock()
+
 	app, err := application.New(logger, pathConfigFile)
 	if err != nil {
 		return err

x-pack/elastic-agent/pkg/artifact/download/fs/downloader.go

@@ -42,7 +42,7 @@ func NewDownloader(config *artifact.Config) *Downloader {
 func (e *Downloader) Download(_ context.Context, programName, version string) (string, error) {
 	// create a destination directory root/program
 	destinationDir := filepath.Join(e.config.TargetDirectory, programName)
-	if err := os.MkdirAll(destinationDir, os.ModeDir); err != nil {
+	if err := os.MkdirAll(destinationDir, 0755); err != nil {
 		return "", errors.New(err, "creating directory for downloaded artifact failed", errors.TypeFilesystem, errors.M(errors.MetaKeyPath, destinationDir))
 	}