Merge pull request #1677 from RichardScothern/tonyhb-fix-s3-gc-error

Move GC into storage package and add tests

registry/root.go

@@ -1,7 +1,14 @@
 package registry
 
 import (
+	"fmt"
+	"os"
+
+	"github.com/docker/distribution/context"
+	"github.com/docker/distribution/registry/storage"
+	"github.com/docker/distribution/registry/storage/driver/factory"
 	"github.com/docker/distribution/version"
+	"github.com/docker/libtrust"
 	"github.com/spf13/cobra"
 )
 
@@ -10,6 +17,7 @@ var showVersion bool
 func init() {
 	RootCmd.AddCommand(ServeCmd)
 	RootCmd.AddCommand(GCCmd)
+	GCCmd.Flags().BoolVarP(&dryRun, "dry-run", "d", false, "do everything except remove the blobs")
 	RootCmd.Flags().BoolVarP(&showVersion, "version", "v", false, "show the version and exit")
 }
 
@@ -26,3 +34,51 @@ var RootCmd = &cobra.Command{
 		cmd.Usage()
 	},
 }
+
+var dryRun bool
+
+// GCCmd is the cobra command that corresponds to the garbage-collect subcommand
+var GCCmd = &cobra.Command{
+	Use:   "garbage-collect <config>",
+	Short: "`garbage-collect` deletes layers not referenced by any manifests",
+	Long:  "`garbage-collect` deletes layers not referenced by any manifests",
+	Run: func(cmd *cobra.Command, args []string) {
+		config, err := resolveConfiguration(args)
+		if err != nil {
+			fmt.Fprintf(os.Stderr, "configuration error: %v\n", err)
+			cmd.Usage()
+			os.Exit(1)
+		}
+
+		driver, err := factory.Create(config.Storage.Type(), config.Storage.Parameters())
+		if err != nil {
+			fmt.Fprintf(os.Stderr, "failed to construct %s driver: %v", config.Storage.Type(), err)
+			os.Exit(1)
+		}
+
+		ctx := context.Background()
+		ctx, err = configureLogging(ctx, config)
+		if err != nil {
+			fmt.Fprintf(os.Stderr, "unable to configure logging with config: %s", err)
+			os.Exit(1)
+		}
+
+		k, err := libtrust.GenerateECP256PrivateKey()
+		if err != nil {
+			fmt.Fprint(os.Stderr, err)
+			os.Exit(1)
+		}
+
+		registry, err := storage.NewRegistry(ctx, driver, storage.DisableSchema1Signatures, storage.Schema1SigningKey(k))
+		if err != nil {
+			fmt.Fprintf(os.Stderr, "failed to construct registry: %v", err)
+			os.Exit(1)
+		}
+
+		err = storage.MarkAndSweep(ctx, driver, registry, dryRun)
+		if err != nil {
+			fmt.Fprintf(os.Stderr, "failed to garbage collect: %v", err)
+			os.Exit(1)
+		}
+	},
+}

registry/storage/garbagecollect.go

@@ -1,8 +1,7 @@
-package registry
+package storage
 
 import (
 	"fmt"
-	"os"
 
 	"github.com/docker/distribution"
 	"github.com/docker/distribution/context"
@@ -10,21 +9,15 @@ import (
 	"github.com/docker/distribution/manifest/schema1"
 	"github.com/docker/distribution/manifest/schema2"
 	"github.com/docker/distribution/reference"
-	"github.com/docker/distribution/registry/storage"
 	"github.com/docker/distribution/registry/storage/driver"
-	"github.com/docker/distribution/registry/storage/driver/factory"
-	"github.com/docker/libtrust"
-	"github.com/spf13/cobra"
 )
 
 func emit(format string, a ...interface{}) {
-	if dryRun {
 	fmt.Printf(format+"\n", a...)
-	}
 }
 
-func markAndSweep(ctx context.Context, storageDriver driver.StorageDriver, registry distribution.Namespace) error {
+// MarkAndSweep performs a mark and sweep of registry data
-
+func MarkAndSweep(ctx context.Context, storageDriver driver.StorageDriver, registry distribution.Namespace, dryRun bool) error {
 	repositoryEnumerator, ok := registry.(distribution.RepositoryEnumerator)
 	if !ok {
 		return fmt.Errorf("unable to convert Namespace to RepositoryEnumerator")
@@ -33,7 +26,9 @@ func markAndSweep(ctx context.Context, storageDriver driver.StorageDriver, regis
 	// mark
 	markSet := make(map[digest.Digest]struct{})
 	err := repositoryEnumerator.Enumerate(ctx, func(repoName string) error {
+		if dryRun {
 			emit(repoName)
+		}
 
 		var err error
 		named, err := reference.ParseNamed(repoName)
@@ -57,7 +52,9 @@ func markAndSweep(ctx context.Context, storageDriver driver.StorageDriver, regis
 
 		err = manifestEnumerator.Enumerate(ctx, func(dgst digest.Digest) error {
 			// Mark the manifest's blob
+			if dryRun {
 				emit("%s: marking manifest %s ", repoName, dgst)
+			}
 			markSet[dgst] = struct{}{}
 
 			manifest, err := manifestService.Get(ctx, dgst)
@@ -68,8 +65,10 @@ func markAndSweep(ctx context.Context, storageDriver driver.StorageDriver, regis
 			descriptors := manifest.References()
 			for _, descriptor := range descriptors {
 				markSet[descriptor.Digest] = struct{}{}
+				if dryRun {
 					emit("%s: marking blob %s", repoName, descriptor.Digest)
 				}
+			}
 
 			switch manifest.(type) {
 			case *schema1.SignedManifest:
@@ -82,13 +81,17 @@ func markAndSweep(ctx context.Context, storageDriver driver.StorageDriver, regis
 					return fmt.Errorf("failed to get signatures for signed manifest: %v", err)
 				}
 				for _, signatureDigest := range signatures {
+					if dryRun {
 						emit("%s: marking signature %s", repoName, signatureDigest)
+					}
 					markSet[signatureDigest] = struct{}{}
 				}
 				break
 			case *schema2.DeserializedManifest:
 				config := manifest.(*schema2.DeserializedManifest).Config
+				if dryRun {
 					emit("%s: marking configuration %s", repoName, config.Digest)
+				}
 				markSet[config.Digest] = struct{}{}
 				break
 			}
@@ -96,6 +99,17 @@ func markAndSweep(ctx context.Context, storageDriver driver.StorageDriver, regis
 			return nil
 		})
 
+		if err != nil {
+			// In certain situations such as unfinished uploads, deleting all
+			// tags in S3 or removing the _manifests folder manually, this
+			// error may be of type PathNotFound.
+			//
+			// In these cases we can continue marking other manifests safely.
+			if _, ok := err.(driver.PathNotFoundError); ok {
+				return nil
+			}
+		}
+
 		return err
 	})
 
@@ -116,13 +130,14 @@ func markAndSweep(ctx context.Context, storageDriver driver.StorageDriver, regis
 	if err != nil {
 		return fmt.Errorf("error enumerating blobs: %v", err)
 	}
-
-	emit("\n%d blobs marked, %d blobs eligible for deletion", len(markSet), len(deleteSet))
-	// Construct vacuum
-	vacuum := storage.NewVacuum(ctx, storageDriver)
-	for dgst := range deleteSet {
-		emit("blob eligible for deletion: %s", dgst)
 	if dryRun {
+		emit("\n%d blobs marked, %d blobs eligible for deletion", len(markSet), len(deleteSet))
+	}
+	// Construct vacuum
+	vacuum := NewVacuum(ctx, storageDriver)
+	for dgst := range deleteSet {
+		if dryRun {
+			emit("blob eligible for deletion: %s", dgst)
 			continue
 		}
 		err = vacuum.RemoveBlob(string(dgst))
@@ -133,55 +148,3 @@ func markAndSweep(ctx context.Context, storageDriver driver.StorageDriver, regis
 
 	return err
 }
-
-func init() {
-	GCCmd.Flags().BoolVarP(&dryRun, "dry-run", "d", false, "do everything except remove the blobs")
-}
-
-var dryRun bool
-
-// GCCmd is the cobra command that corresponds to the garbage-collect subcommand
-var GCCmd = &cobra.Command{
-	Use:   "garbage-collect <config>",
-	Short: "`garbage-collect` deletes layers not referenced by any manifests",
-	Long:  "`garbage-collect` deletes layers not referenced by any manifests",
-	Run: func(cmd *cobra.Command, args []string) {
-		config, err := resolveConfiguration(args)
-		if err != nil {
-			fmt.Fprintf(os.Stderr, "configuration error: %v\n", err)
-			cmd.Usage()
-			os.Exit(1)
-		}
-
-		driver, err := factory.Create(config.Storage.Type(), config.Storage.Parameters())
-		if err != nil {
-			fmt.Fprintf(os.Stderr, "failed to construct %s driver: %v", config.Storage.Type(), err)
-			os.Exit(1)
-		}
-
-		ctx := context.Background()
-		ctx, err = configureLogging(ctx, config)
-		if err != nil {
-			fmt.Fprintf(os.Stderr, "unable to configure logging with config: %s", err)
-			os.Exit(1)
-		}
-
-		k, err := libtrust.GenerateECP256PrivateKey()
-		if err != nil {
-			fmt.Fprint(os.Stderr, err)
-			os.Exit(1)
-		}
-
-		registry, err := storage.NewRegistry(ctx, driver, storage.DisableSchema1Signatures, storage.Schema1SigningKey(k))
-		if err != nil {
-			fmt.Fprintf(os.Stderr, "failed to construct registry: %v", err)
-			os.Exit(1)
-		}
-
-		err = markAndSweep(ctx, driver, registry)
-		if err != nil {
-			fmt.Fprintf(os.Stderr, "failed to garbage collect: %v", err)
-			os.Exit(1)
-		}
-	},
-}

registry/storage/garbagecollect_test.go

@@ -1,14 +1,14 @@
-package registry
+package storage
 
 import (
 	"io"
+	"path"
 	"testing"
 
 	"github.com/docker/distribution"
 	"github.com/docker/distribution/context"
 	"github.com/docker/distribution/digest"
 	"github.com/docker/distribution/reference"
-	"github.com/docker/distribution/registry/storage"
 	"github.com/docker/distribution/registry/storage/driver"
 	"github.com/docker/distribution/registry/storage/driver/inmemory"
 	"github.com/docker/distribution/testutil"
@@ -22,7 +22,7 @@ type image struct {
 
 func createRegistry(t *testing.T, driver driver.StorageDriver) distribution.Namespace {
 	ctx := context.Background()
-	registry, err := storage.NewRegistry(ctx, driver, storage.EnableDelete)
+	registry, err := NewRegistry(ctx, driver, EnableDelete)
 	if err != nil {
 		t.Fatalf("Failed to construct namespace")
 	}
@@ -161,7 +161,7 @@ func TestNoDeletionNoEffect(t *testing.T) {
 	}
 
 	// Run GC
-	err = markAndSweep(context.Background(), inmemoryDriver, registry)
+	err = MarkAndSweep(context.Background(), inmemoryDriver, registry, false)
 	if err != nil {
 		t.Fatalf("Failed mark and sweep: %v", err)
 	}
@@ -177,6 +177,37 @@ func TestNoDeletionNoEffect(t *testing.T) {
 	}
 }
 
+func TestGCWithMissingManifests(t *testing.T) {
+	ctx := context.Background()
+	d := inmemory.New()
+
+	registry := createRegistry(t, d)
+	repo := makeRepository(t, registry, "testrepo")
+	uploadRandomSchema1Image(t, repo)
+
+	// Simulate a missing _manifests directory
+	revPath, err := pathFor(manifestRevisionsPathSpec{"testrepo"})
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	_manifestsPath := path.Dir(revPath)
+	err = d.Delete(ctx, _manifestsPath)
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	err = MarkAndSweep(context.Background(), d, registry, false)
+	if err != nil {
+		t.Fatalf("Failed mark and sweep: %v", err)
+	}
+
+	blobs := allBlobs(t, registry)
+	if len(blobs) > 0 {
+		t.Errorf("unexpected blobs after gc")
+	}
+}
+
 func TestDeletionHasEffect(t *testing.T) {
 	ctx := context.Background()
 	inmemoryDriver := inmemory.New()
@@ -193,7 +224,7 @@ func TestDeletionHasEffect(t *testing.T) {
 	manifests.Delete(ctx, image3.manifestDigest)
 
 	// Run GC
-	err = markAndSweep(context.Background(), inmemoryDriver, registry)
+	err = MarkAndSweep(context.Background(), inmemoryDriver, registry, false)
 	if err != nil {
 		t.Fatalf("Failed mark and sweep: %v", err)
 	}
@@ -327,7 +358,7 @@ func TestOrphanBlobDeleted(t *testing.T) {
 	uploadRandomSchema2Image(t, repo)
 
 	// Run GC
-	err = markAndSweep(context.Background(), inmemoryDriver, registry)
+	err = MarkAndSweep(context.Background(), inmemoryDriver, registry, false)
 	if err != nil {
 		t.Fatalf("Failed mark and sweep: %v", err)
 	}