fix(server): Use uncompressed tarball hashes in image config

Docker expects hashes of compressed tarballs in the manifest (as these
are used to fetch from the content-addressable layer store), but for
some reason it expects hashes in the configuration layer to be of
uncompressed tarballs.

To achieve this an additional SHA256 hash is calculcated while
creating the layer tarballs, but before passing them to the gzip
writer.

In the current constellation the symlink layer is first compressed and
then decompressed again to calculate its hash. This can be refactored
in a future change.

4 files changed, 42 insertions, 16 deletions

diff --git a/tools/nixery/build-image/build-image.nix b/tools/nixery/build-image/build-image.nix
index b78ee66264..ab785006a9 100644
--- a/tools/nixery/build-image/build-image.nix
+++ b/tools/nixery/build-image/build-image.nix
@@ -137,11 +137,14 @@ let
   symlinkLayerMeta = fromJSON (readFile (runCommand "symlink-layer-meta.json" {
     buildInputs = with pkgs; [ coreutils jq openssl ];
   }''
-    layerSha256=$(sha256sum ${symlinkLayer} | cut -d ' ' -f1)
+    gzipHash=$(sha256sum ${symlinkLayer} | cut -d ' ' -f1)
+    tarHash=$(cat ${symlinkLayer} | gzip -d | sha256sum | cut -d ' ' -f1)
     layerSize=$(stat --printf '%s' ${symlinkLayer})
 
-    jq -n -c --arg sha256 $layerSha256 --arg size $layerSize --arg path ${symlinkLayer} \
-      '{ size: ($size | tonumber), sha256: $sha256, path: $path }' >> $out
+    jq -n -c --arg gzipHash $gzipHash --arg tarHash $tarHash --arg size $layerSize \
+      --arg path ${symlinkLayer} \
+      '{ size: ($size | tonumber), tarHash: $tarHash, gzipHash: $gzipHash, path: $path }' \
+      >> $out
   ''));
 
   # Final output structure returned to Nixery if the build succeeded
diff --git a/tools/nixery/server/builder/archive.go b/tools/nixery/server/builder/archive.go
index 55b3c2a8b7..a3fb99882f 100644
--- a/tools/nixery/server/builder/archive.go
+++ b/tools/nixery/server/builder/archive.go
@@ -10,6 +10,8 @@ package builder
 import (
 	"archive/tar"
 	"compress/gzip"
+	"crypto/sha256"
+	"fmt"
 	"io"
 	"os"
 	"path/filepath"
@@ -19,26 +21,31 @@ import (
 
 // Create a new compressed tarball from each of the paths in the list
 // and write it to the supplied writer.
-func packStorePaths(l *layers.Layer, w io.Writer) error {
+//
+// The uncompressed tarball is hashed because image manifests must
+// contain both the hashes of compressed and uncompressed layers.
+func packStorePaths(l *layers.Layer, w io.Writer) (string, error) {
+	shasum := sha256.New()
 	gz := gzip.NewWriter(w)
-	t := tar.NewWriter(gz)
+	multi := io.MultiWriter(shasum, gz)
+	t := tar.NewWriter(multi)
 
 	for _, path := range l.Contents {
 		err := filepath.Walk(path, tarStorePath(t))
 		if err != nil {
-			return err
+			return "", err
 		}
 	}
 
 	if err := t.Close(); err != nil {
-		return err
+		return "", err
 	}
 
 	if err := gz.Close(); err != nil {
-		return err
+		return "", err
 	}
 
-	return nil
+	return fmt.Sprintf("sha256:%x", shasum.Sum([]byte{})), nil
 }
 
 func tarStorePath(w *tar.Writer) filepath.WalkFunc {
diff --git a/tools/nixery/server/builder/builder.go b/tools/nixery/server/builder/builder.go
index 748ff5f67d..39befd0fb8 100644
--- a/tools/nixery/server/builder/builder.go
+++ b/tools/nixery/server/builder/builder.go
@@ -117,9 +117,10 @@ type ImageResult struct {
 	// These fields are populated in case of success
 	Graph        layers.RuntimeGraph `json:"runtimeGraph"`
 	SymlinkLayer struct {
-		Size   int    `json:"size"`
-		SHA256 string `json:"sha256"`
-		Path   string `json:"path"`
+		Size     int    `json:"size"`
+		TarHash  string `json:"tarHash"`
+		GzipHash string `json:"gzipHash"`
+		Path     string `json:"path"`
 	} `json:"symlinkLayer"`
 }
 
@@ -269,8 +270,18 @@ func prepareLayers(ctx context.Context, s *State, image *Image, result *ImageRes
 			entries = append(entries, *entry)
 		} else {
 			lh := l.Hash()
+
+			// While packing store paths, the SHA sum of
+			// the uncompressed layer is computed and
+			// written to `tarhash`.
+			//
+			// TODO(tazjin): Refactor this to make the
+			// flow of data cleaner.
+			var tarhash string
 			lw := func(w io.Writer) error {
-				return packStorePaths(&l, w)
+				var err error
+				tarhash, err = packStorePaths(&l, w)
+				return err
 			}
 
 			entry, err := uploadHashLayer(ctx, s, lh, lw)
@@ -278,6 +289,7 @@ func prepareLayers(ctx context.Context, s *State, image *Image, result *ImageRes
 				return nil, err
 			}
 			entry.MergeRating = l.MergeRating
+			entry.TarHash = tarhash
 
 			var pkgs []string
 			for _, p := range l.Contents {
@@ -287,6 +299,7 @@ func prepareLayers(ctx context.Context, s *State, image *Image, result *ImageRes
 			log.WithFields(log.Fields{
 				"layer":    lh,
 				"packages": pkgs,
+				"tarhash":  tarhash,
 			}).Info("created image layer")
 
 			go cacheLayer(ctx, s, l.Hash(), *entry)
@@ -296,7 +309,7 @@ func prepareLayers(ctx context.Context, s *State, image *Image, result *ImageRes
 
 	// Symlink layer (built in the first Nix build) needs to be
 	// included here manually:
-	slkey := result.SymlinkLayer.SHA256
+	slkey := result.SymlinkLayer.GzipHash
 	entry, err := uploadHashLayer(ctx, s, slkey, func(w io.Writer) error {
 		f, err := os.Open(result.SymlinkLayer.Path)
 		if err != nil {
@@ -318,6 +331,7 @@ func prepareLayers(ctx context.Context, s *State, image *Image, result *ImageRes
 		return nil, err
 	}
 
+	entry.TarHash = "sha256:" + result.SymlinkLayer.TarHash
 	go cacheLayer(ctx, s, slkey, *entry)
 	entries = append(entries, *entry)
 
diff --git a/tools/nixery/server/manifest/manifest.go b/tools/nixery/server/manifest/manifest.go
index 8e65fa223b..8ad8282395 100644
--- a/tools/nixery/server/manifest/manifest.go
+++ b/tools/nixery/server/manifest/manifest.go
@@ -29,9 +29,10 @@ type Entry struct {
 	Size      int64  `json:"size"`
 	Digest    string `json:"digest"`
 
-	// This field is internal to Nixery and not part of the
+	// These fields are internal to Nixery and not part of the
 	// serialised entry.
 	MergeRating uint64 `json:"-"`
+	TarHash     string `json:",omitempty"`
 }
 
 type manifest struct {
@@ -102,9 +103,10 @@ func Manifest(layers []Entry) (json.RawMessage, ConfigLayer) {
 
 	hashes := make([]string, len(layers))
 	for i, l := range layers {
+		hashes[i] = l.TarHash
 		l.MediaType = layerType
+		l.TarHash = ""
 		layers[i] = l
-		hashes[i] = l.Digest
 	}
 
 	c := configLayer(hashes)