From cb9e773a3ec61d3e807e123cc816c424283d0647 Mon Sep 17 00:00:00 2001
From: Ramkumar Chinchani <rchincha@cisco.com>
Date: Wed, 27 May 2020 14:27:35 -0700
Subject: [PATCH] dedupe: record relative path for cache entries

In a production use case we found that the actual rootdir can be moved.
Currently, cache entries for dedupe record the full blob path which
doesn't work in the move use case.

Only for dedupe cache entries, record relative blob paths.
---
 pkg/storage/cache.go      | 32 +++++++++++++++++++++++---------
 pkg/storage/cache_test.go |  3 ++-
 pkg/storage/storage.go    |  2 ++
 3 files changed, 27 insertions(+), 10 deletions(-)

diff --git a/pkg/storage/cache.go b/pkg/storage/cache.go
index 1c77f76b..bd863293 100644
--- a/pkg/storage/cache.go
+++ b/pkg/storage/cache.go
@@ -2,6 +2,7 @@ package storage
 
 import (
 	"path"
+	"path/filepath"
 	"strings"
 
 	"github.com/anuvu/zot/errors"
@@ -14,8 +15,9 @@ const (
 )
 
 type Cache struct {
-	db  *bbolt.DB
-	log zlog.Logger
+	rootDir string
+	db      *bbolt.DB
+	log     zlog.Logger
 }
 
 // Blob is a blob record
@@ -45,10 +47,16 @@ func NewCache(rootDir string, name string, log zlog.Logger) *Cache {
 		return nil
 	}
 
-	return &Cache{db: db, log: log}
+	return &Cache{rootDir: rootDir, db: db, log: log}
 }
 
 func (c *Cache) PutBlob(digest string, path string) error {
+	// use only relative (to rootDir) paths on blobs
+	relp, err := filepath.Rel(c.rootDir, path)
+	if err != nil {
+		c.log.Error().Err(err).Str("path", path).Msg("unable to get relative path")
+	}
+
 	if err := c.db.Update(func(tx *bbolt.Tx) error {
 		root := tx.Bucket([]byte(BlobsCache))
 		if root == nil {
@@ -63,8 +71,8 @@ func (c *Cache) PutBlob(digest string, path string) error {
 			c.log.Error().Err(err).Str("bucket", digest).Msg("unable to create a bucket")
 			return err
 		}
-		if err := b.Put([]byte(path), nil); err != nil {
-			c.log.Error().Err(err).Str("bucket", digest).Str("value", path).Msg("unable to put record")
+		if err := b.Put([]byte(relp), nil); err != nil {
+			c.log.Error().Err(err).Str("bucket", digest).Str("value", relp).Msg("unable to put record")
 			return err
 		}
 		return nil
@@ -135,6 +143,12 @@ func (c *Cache) HasBlob(digest string, blob string) bool {
 }
 
 func (c *Cache) DeleteBlob(digest string, path string) error {
+	// use only relative (to rootDir) paths on blobs
+	relp, err := filepath.Rel(c.rootDir, path)
+	if err != nil {
+		c.log.Error().Err(err).Str("path", path).Msg("unable to get relative path")
+	}
+
 	if err := c.db.Update(func(tx *bbolt.Tx) error {
 		root := tx.Bucket([]byte(BlobsCache))
 		if root == nil {
@@ -149,8 +163,8 @@ func (c *Cache) DeleteBlob(digest string, path string) error {
 			return errors.ErrCacheMiss
 		}
 
-		if err := b.Delete([]byte(path)); err != nil {
-			c.log.Error().Err(err).Str("digest", digest).Str("path", path).Msg("unable to delete")
+		if err := b.Delete([]byte(relp)); err != nil {
+			c.log.Error().Err(err).Str("digest", digest).Str("path", relp).Msg("unable to delete")
 			return err
 		}
 
@@ -158,9 +172,9 @@ func (c *Cache) DeleteBlob(digest string, path string) error {
 		k, _ := cur.First()
 
 		if k == nil {
-			c.log.Debug().Str("digest", digest).Str("path", path).Msg("deleting empty bucket")
+			c.log.Debug().Str("digest", digest).Str("path", relp).Msg("deleting empty bucket")
 			if err := root.DeleteBucket([]byte(digest)); err != nil {
-				c.log.Error().Err(err).Str("digest", digest).Str("path", path).Msg("unable to delete")
+				c.log.Error().Err(err).Str("digest", digest).Str("path", relp).Msg("unable to delete")
 				return err
 			}
 		}
diff --git a/pkg/storage/cache_test.go b/pkg/storage/cache_test.go
index 05ea6c61..0e49dc4e 100644
--- a/pkg/storage/cache_test.go
+++ b/pkg/storage/cache_test.go
@@ -3,6 +3,7 @@ package storage_test
 import (
 	"io/ioutil"
 	"os"
+	"path"
 	"testing"
 
 	"github.com/anuvu/zot/errors"
@@ -33,7 +34,7 @@ func TestCache(t *testing.T) {
 		b := c.HasBlob("key", "value")
 		So(b, ShouldBeFalse)
 
-		err = c.PutBlob("key", "value")
+		err = c.PutBlob("key", path.Join(dir, "value"))
 		So(err, ShouldBeNil)
 
 		b = c.HasBlob("key", "value")
diff --git a/pkg/storage/storage.go b/pkg/storage/storage.go
index 71602f7e..e0b61a5f 100644
--- a/pkg/storage/storage.go
+++ b/pkg/storage/storage.go
@@ -849,6 +849,8 @@ retry:
 		}
 		is.log.Debug().Str("src", src).Str("dst", dst).Msg("dedupe: rename")
 	} else {
+		dstRecord = path.Join(is.rootDir, dstRecord)
+
 		dstRecordFi, err := os.Stat(dstRecord)
 		if err != nil {
 			is.log.Error().Err(err).Str("blobPath", dstRecord).Msg("dedupe: unable to stat")