Refactor the build loop. (#385)

This change refactors the build loop a bit to make cache optimization easier in the future. Some notable changes:

The special casing around volume snapshots is removed. Every volume is added to the snapshotFiles list for every command that will snapshot anyway.
Snapshot saving was extracted to a sub-function
The decision on whether or not to snapshot was extracted

pkg/commands/volume.go

@@ -30,8 +30,7 @@ import (
 
 type VolumeCommand struct {
 	BaseCommand
-	cmd           *instructions.VolumeCommand
+	cmd *instructions.VolumeCommand
-	snapshotFiles []string
 }
 
 func (v *VolumeCommand) ExecuteCommand(config *v1.Config, buildArgs *dockerfile.BuildArgs) error {
@@ -57,7 +56,6 @@ func (v *VolumeCommand) ExecuteCommand(config *v1.Config, buildArgs *dockerfile.
 		// Only create and snapshot the dir if it didn't exist already
 		if _, err := os.Stat(volume); os.IsNotExist(err) {
 			logrus.Infof("Creating directory %s", volume)
-			v.snapshotFiles = append(v.snapshotFiles, volume)
 			if err := os.MkdirAll(volume, 0755); err != nil {
 				return fmt.Errorf("Could not create directory for volume %s: %s", volume, err)
 			}
@@ -69,7 +67,7 @@ func (v *VolumeCommand) ExecuteCommand(config *v1.Config, buildArgs *dockerfile.
 }
 
 func (v *VolumeCommand) FilesToSnapshot() []string {
-	return v.snapshotFiles
+	return []string{}
 }
 
 func (v *VolumeCommand) String() string {

pkg/commands/volume_test.go

@@ -43,7 +43,6 @@ func TestUpdateVolume(t *testing.T) {
 		cmd: &instructions.VolumeCommand{
 			Volumes: volumes,
 		},
-		snapshotFiles: []string{},
 	}
 
 	expectedVolumes := map[string]struct{}{

pkg/constants/constants.go

@@ -62,9 +62,6 @@ const (
 	// Docker command names
 	Cmd        = "cmd"
 	Entrypoint = "entrypoint"
-
-	// VolumeCmdName is the name of the volume command
-	VolumeCmdName = "volume"
 )
 
 // KanikoBuildFiles is the list of files required to build kaniko

pkg/executor/build.go

@@ -49,6 +49,7 @@ type stageBuilder struct {
 	cf              *v1.ConfigFile
 	snapshotter     *snapshot.Snapshotter
 	baseImageDigest string
+	opts            *config.KanikoOptions
 }
 
 // newStageBuilder returns a new type stageBuilder which contains all the information required to build the stage
@@ -81,6 +82,7 @@ func newStageBuilder(opts *config.KanikoOptions, stage config.KanikoStage) (*sta
 		cf:              imageConfig,
 		snapshotter:     snapshotter,
 		baseImageDigest: digest.String(),
+		opts:            opts,
 	}, nil
 }
 
@@ -111,7 +113,7 @@ func (s *stageBuilder) extractCachedLayer(layer v1.Image, createdBy string) erro
 	return err
 }
 
-func (s *stageBuilder) build(opts *config.KanikoOptions) error {
+func (s *stageBuilder) build() error {
 	// Unpack file system to root
 	if _, err := util.GetFSFromImage(constants.RootDir, s.image); err != nil {
 		return err
@@ -120,23 +122,26 @@ func (s *stageBuilder) build(opts *config.KanikoOptions) error {
 	if err := s.snapshotter.Init(); err != nil {
 		return err
 	}
-	var volumes []string
 
 	// Set the initial cache key to be the base image digest, the build args and the SrcContext.
 	compositeKey := NewCompositeCache(s.baseImageDigest)
-	contextHash, err := HashDir(opts.SrcContext)
+	contextHash, err := HashDir(s.opts.SrcContext)
 	if err != nil {
 		return err
 	}
-	compositeKey.AddKey(opts.BuildArgs...)
+	compositeKey.AddKey(s.opts.BuildArgs...)
 
-	args := dockerfile.NewBuildArgs(opts.BuildArgs)
+	cmds := []commands.DockerCommand{}
-	for index, cmd := range s.stage.Commands {
+	for _, cmd := range s.stage.Commands {
-		finalCmd := index == len(s.stage.Commands)-1
+		command, err := commands.GetCommand(cmd, s.opts.SrcContext)
-		command, err := commands.GetCommand(cmd, opts.SrcContext)
 		if err != nil {
 			return err
 		}
+		cmds = append(cmds, command)
+	}
+
+	args := dockerfile.NewBuildArgs(s.opts.BuildArgs)
+	for index, command := range cmds {
 		if command == nil {
 			continue
 		}
@@ -153,8 +158,8 @@ func (s *stageBuilder) build(opts *config.KanikoOptions) error {
 			return err
 		}
 
-		if command.CacheCommand() && opts.Cache {
+		if command.CacheCommand() && s.opts.Cache {
-			image, err := cache.RetrieveLayer(opts, ck)
+			image, err := cache.RetrieveLayer(s.opts, ck)
 			if err == nil {
 				if err := s.extractCachedLayer(image, command.String()); err != nil {
 					return errors.Wrap(err, "extracting cached layer")
@@ -163,84 +168,94 @@ func (s *stageBuilder) build(opts *config.KanikoOptions) error {
 			}
 			logrus.Info("No cached layer found, executing command...")
 		}
+
 		if err := command.ExecuteCommand(&s.cf.Config, args); err != nil {
 			return err
 		}
 		files := command.FilesToSnapshot()
-		if cmd.Name() == constants.VolumeCmdName {
-			volumes = append(volumes, files...)
-			continue
-		}
 		var contents []byte
 
-		// If this is an intermediate stage, we only snapshot for the last command and we
+		if !s.shouldTakeSnapshot(index, files) {
-		// want to snapshot the entire filesystem since we aren't tracking what was changed
-		// by previous commands.
-		if !s.stage.Final {
-			if finalCmd {
-				contents, err = s.snapshotter.TakeSnapshotFS()
-			}
-		} else {
-			// If we are in single snapshot mode, we only take a snapshot once, after all
-			// commands have completed.
-			if opts.SingleSnapshot {
-				if finalCmd {
-					contents, err = s.snapshotter.TakeSnapshotFS()
-				}
-			} else {
-				// Otherwise, in the final stage we take a snapshot at each command. If we know
-				// the files that were changed, we'll snapshot those explicitly, otherwise we'll
-				// check if anything in the filesystem changed.
-				if files != nil {
-					if len(files) > 0 {
-						files = append(files, volumes...)
-						volumes = []string{}
-					}
-					contents, err = s.snapshotter.TakeSnapshot(files)
-				} else {
-					contents, err = s.snapshotter.TakeSnapshotFS()
-					volumes = []string{}
-				}
-			}
-		}
-		if err != nil {
-			return fmt.Errorf("Error taking snapshot of files for command %s: %s", command, err)
-		}
-
-		if contents == nil {
-			logrus.Info("No files were changed, appending empty layer to config. No layer added to image.")
 			continue
 		}
-		// Append the layer to the image
+
-		opener := func() (io.ReadCloser, error) {
+		if files == nil || s.opts.SingleSnapshot {
-			return ioutil.NopCloser(bytes.NewReader(contents)), nil
+			contents, err = s.snapshotter.TakeSnapshotFS()
+		} else {
+			// Volumes are very weird. They get created in their command, but snapshotted in the next one.
+			// Add them to the list of files to snapshot.
+			for v := range s.cf.Config.Volumes {
+				files = append(files, v)
+			}
+			contents, err = s.snapshotter.TakeSnapshot(files)
 		}
-		layer, err := tarball.LayerFromOpener(opener)
 		if err != nil {
 			return err
 		}
-		// Push layer to cache now along with new config file
+		if err := s.saveSnapshot(command, ck, contents); err != nil {
-		if command.CacheCommand() && opts.Cache {
-			if err := pushLayerToCache(opts, ck, layer, command.String()); err != nil {
-				return err
-			}
-		}
-		s.image, err = mutate.Append(s.image,
-			mutate.Addendum{
-				Layer: layer,
-				History: v1.History{
-					Author:    constants.Author,
-					CreatedBy: command.String(),
-				},
-			},
-		)
-		if err != nil {
 			return err
 		}
 	}
 	return nil
 }
 
+func (s *stageBuilder) shouldTakeSnapshot(index int, files []string) bool {
+	isLastCommand := index == len(s.stage.Commands)-1
+
+	// We only snapshot the very end of intermediate stages.
+	if !s.stage.Final {
+		return isLastCommand
+	}
+
+	// We only snapshot the very end with single snapshot mode on.
+	if s.opts.SingleSnapshot {
+		return isLastCommand
+	}
+
+	// nil means snapshot everything.
+	if files == nil {
+		return true
+	}
+
+	// Don't snapshot an empty list.
+	if len(files) == 0 {
+		return false
+	}
+	return true
+}
+
+func (s *stageBuilder) saveSnapshot(command commands.DockerCommand, ck string, contents []byte) error {
+	if contents == nil {
+		logrus.Info("No files were changed, appending empty layer to config. No layer added to image.")
+		return nil
+	}
+	// Append the layer to the image
+	opener := func() (io.ReadCloser, error) {
+		return ioutil.NopCloser(bytes.NewReader(contents)), nil
+	}
+	layer, err := tarball.LayerFromOpener(opener)
+	if err != nil {
+		return err
+	}
+	// Push layer to cache now along with new config file
+	if command.CacheCommand() && s.opts.Cache {
+		if err := pushLayerToCache(s.opts, ck, layer, command.String()); err != nil {
+			return err
+		}
+	}
+	s.image, err = mutate.Append(s.image,
+		mutate.Addendum{
+			Layer: layer,
+			History: v1.History{
+				Author:    constants.Author,
+				CreatedBy: command.String(),
+			},
+		},
+	)
+	return err
+
+}
+
 // DoBuild executes building the Dockerfile
 func DoBuild(opts *config.KanikoOptions) (v1.Image, error) {
 	// Parse dockerfile and unpack base image to root
@@ -253,7 +268,7 @@ func DoBuild(opts *config.KanikoOptions) (v1.Image, error) {
 		if err != nil {
 			return nil, errors.Wrap(err, fmt.Sprintf("getting stage builder for stage %d", index))
 		}
-		if err := sb.build(opts); err != nil {
+		if err := sb.build(); err != nil {
 			return nil, errors.Wrap(err, "error building stage")
 		}
 		reviewConfig(stage, &sb.cf.Config)