Use tar for intermediate backup storage

During the backup we create an intermediate copy of two directories:
builds and uploads. Instead of creating many small files with 'cp
-r', we now use tar (and fast gzip) to create single intermediate
files. This saves on disk IO and disk space while creating a backup.

4 files changed, 45 insertions, 59 deletions

diff --git a/lib/backup/builds.rb b/lib/backup/builds.rb
index 6f56f680bb9..d269f8e260c 100644
--- a/lib/backup/builds.rb
+++ b/lib/backup/builds.rb
@@ -1,34 +1,7 @@
 module Backup
-  class Builds
-    attr_reader :app_builds_dir, :backup_builds_dir, :backup_dir
-
+  class Builds < Files
     def initialize
-      @app_builds_dir = Settings.gitlab_ci.builds_path
-      @backup_dir = Gitlab.config.backup.path
-      @backup_builds_dir = File.join(Gitlab.config.backup.path, 'builds')
-    end
-
-    # Copy builds from builds directory to backup/builds
-    def dump
-      FileUtils.rm_rf(backup_builds_dir)
-      # Ensure the parent dir of backup_builds_dir exists
-      FileUtils.mkdir_p(Gitlab.config.backup.path)
-      # Fail if somebody raced to create backup_builds_dir before us
-      FileUtils.mkdir(backup_builds_dir, mode: 0700)
-      FileUtils.cp_r(app_builds_dir, backup_dir)
-    end
-
-    def restore
-      backup_existing_builds_dir
-
-      FileUtils.cp_r(backup_builds_dir, app_builds_dir)
-    end
-
-    def backup_existing_builds_dir
-      timestamped_builds_path = File.join(app_builds_dir, '..', "builds.#{Time.now.to_i}")
-      if File.exists?(app_builds_dir)
-        FileUtils.mv(app_builds_dir, File.expand_path(timestamped_builds_path))
-      end
+      super(Settings.gitlab_ci.builds_path)
     end
   end
 end
diff --git a/lib/backup/files.rb b/lib/backup/files.rb
new file mode 100644
index 00000000000..d0a6e8f27be
--- /dev/null
+++ b/lib/backup/files.rb
@@ -0,0 +1,39 @@
+require 'open3'
+
+module Backup
+  class Files
+    attr_reader :name, :app_files_dir, :backup_tarball, :backup_dir, :files_parent_dir
+
+    def initialize(app_files_dir)
+      @app_files_dir = File.realpath(app_files_dir)
+      @name = File.basename(app_files_dir)
+      @files_parent_dir = File.realpath(File.join(@app_files_dir, '..'))
+      @backup_dir = Gitlab.config.backup.path
+      @backup_tarball = File.join(@backup_dir, name + '.tar.gz')
+    end
+
+    # Copy files from public/files to backup/files
+    def dump
+      FileUtils.mkdir_p(Gitlab.config.backup.path)
+      run_pipeline!([%W(tar -C #{files_parent_dir} -cf - #{name}), %W(gzip -c -1)], out: [backup_tarball, 'w', 0600])
+    end
+
+    def restore
+      backup_existing_files_dir
+
+      run_pipeline!([%W(gzip -cd), %W(tar -C #{files_parent_dir} -xf -)], in: backup_tarball)
+    end
+
+    def backup_existing_files_dir
+      timestamped_files_path = File.join(files_parent_dir, "#{name}.#{Time.now.to_i}")
+      if File.exists?(app_files_dir)
+        FileUtils.mv(app_files_dir, File.expand_path(timestamped_files_path))
+      end
+    end
+
+    def run_pipeline!(cmd_list, options={})
+      status_list = Open3.pipeline(*cmd_list, options)
+      abort 'Backup failed' unless status_list.compact.all?(&:success?)
+    end
+  end
+end
diff --git a/lib/backup/manager.rb b/lib/backup/manager.rb
index 5c42f25f4a2..f011fd03de0 100644
--- a/lib/backup/manager.rb
+++ b/lib/backup/manager.rb
@@ -150,11 +150,11 @@ module Backup
     private
 
     def backup_contents
-      folders_to_backup + ["backup_information.yml"]
+      folders_to_backup + ["uploads.tar.gz", "builds.tar.gz", "backup_information.yml"]
     end
 
     def folders_to_backup
-      folders = %w{repositories db uploads builds}
+      folders = %w{repositories db}
 
       if ENV["SKIP"]
         return folders.reject{ |folder| ENV["SKIP"].include?(folder) }
diff --git a/lib/backup/uploads.rb b/lib/backup/uploads.rb
index 1f9626644e6..7c0838cc8b7 100644
--- a/lib/backup/uploads.rb
+++ b/lib/backup/uploads.rb
@@ -1,34 +1,8 @@
 module Backup
-  class Uploads
-    attr_reader :app_uploads_dir, :backup_uploads_dir, :backup_dir
+  class Uploads < Files
 
     def initialize
-      @app_uploads_dir = File.realpath(Rails.root.join('public', 'uploads'))
-      @backup_dir = Gitlab.config.backup.path
-      @backup_uploads_dir = File.join(Gitlab.config.backup.path, 'uploads')
-    end
-
-    # Copy uploads from public/uploads to backup/uploads
-    def dump
-      FileUtils.rm_rf(backup_uploads_dir)
-      # Ensure the parent dir of backup_uploads_dir exists
-      FileUtils.mkdir_p(Gitlab.config.backup.path)
-      # Fail if somebody raced to create backup_uploads_dir before us
-      FileUtils.mkdir(backup_uploads_dir, mode: 0700)
-      FileUtils.cp_r(app_uploads_dir, backup_dir)
-    end
-
-    def restore
-      backup_existing_uploads_dir
-
-      FileUtils.cp_r(backup_uploads_dir, app_uploads_dir)
-    end
-
-    def backup_existing_uploads_dir
-      timestamped_uploads_path = File.join(app_uploads_dir, '..', "uploads.#{Time.now.to_i}")
-      if File.exists?(app_uploads_dir)
-        FileUtils.mv(app_uploads_dir, File.expand_path(timestamped_uploads_path))
-      end
+      super(Rails.root.join('public/uploads'))
     end
   end
 end