object-storage: use temp file optimization for pkl

Override OpenStack Swift's swift.common.utils.write_pickle with our own
implementation that uses the GlusterFS temporary file operation. A file name
'.<FILENAME>.<RANDOM>' will hash to the same GlusterFS node as a file named
'<FILENAME>', those avoiding a move of the file on a rename.

This is part of the work needed to address BZ 876660
(https://bugzilla.redhat.com/show_bug.cgi?id=876660).

Change-Id: I1cb9f97f289ab2ca76ec9221366df74de08268bb
BUG: 876660
Signed-off-by: Peter Portante <peter.portante@redhat.com>
Reviewed-on: http://review.gluster.org/4224
Reviewed-by: Kaleb KEITHLEY <kkeithle@redhat.com>
Reviewed-by: Mohammed Junaid <junaid@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>

3 files changed, 73 insertions, 10 deletions

diff --git a/ufo/gluster/swift/common/utils.py b/ufo/gluster/swift/common/utils.py
index 56376f8ee..d35abe582 100644
--- a/ufo/gluster/swift/common/utils.py
+++ b/ufo/gluster/swift/common/utils.py
@@ -17,6 +17,7 @@ import logging
 import os
 import errno
 import xattr
+import random
 from hashlib import md5
 import cPickle as pickle
 from ConfigParser import ConfigParser, NoSectionError, NoOptionError
@@ -455,3 +456,33 @@ def create_container_metadata(cont_path, memcache=None):
 def create_account_metadata(acc_path, memcache=None):
     metadata = get_account_metadata(acc_path, memcache)
     return restore_metadata(acc_path, metadata)
+
+def write_pickle(obj, dest, tmp=None, pickle_protocol=0):
+    """
+    Ensure that a pickle file gets written to disk.  The file is first written
+    to a tmp file location in the destination directory path, ensured it is
+    synced to disk, then moved to its final destination name.
+
+    This version takes advantage of Gluster's dot-prefix-dot-suffix naming
+    where the a file named ".thefile.name.9a7aasv" is hashed to the same
+    Gluster node as "thefile.name". This ensures the renaming of a temp file
+    once written does not move it to another Gluster node.
+
+    :param obj: python object to be pickled
+    :param dest: path of final destination file
+    :param tmp: path to tmp to use, defaults to None (ignored)
+    :param pickle_protocol: protocol to pickle the obj with, defaults to 0
+    """
+    dirname = os.path.dirname(dest)
+    basename = os.path.basename(dest)
+    tmpname = '.' + basename + '.' + md5(basename + str(random.random())).hexdigest()
+    tmppath = os.path.join(dirname, tmpname)
+    with open(tmppath, 'wb') as fo:
+        pickle.dump(obj, fo, pickle_protocol)
+        fo.flush()
+        os.fsync(fo)
+    os.rename(tmppath, dest)
+
+# Over-ride Swift's utils.write_pickle with ours
+import swift.common.utils
+swift.common.utils.write_pickle = write_pickle
diff --git a/ufo/gluster/swift/obj/server.py b/ufo/gluster/swift/obj/server.py
index 43cdd8890..1c2b6cb1d 100644
--- a/ufo/gluster/swift/obj/server.py
+++ b/ufo/gluster/swift/obj/server.py
@@ -18,6 +18,7 @@
 # Simply importing this monkey patches the constraint handling to fit our
 # needs
 import gluster.swift.common.constraints
+import gluster.swift.common.utils
 
 from swift.obj import server
 from gluster.swift.common.DiskFile import Gluster_DiskFile
diff --git a/ufo/test/unit/common/test_utils.py b/ufo/test/unit/common/test_utils.py
index 21acd8e90..446939963 100644
--- a/ufo/test/unit/common/test_utils.py
+++ b/ufo/test/unit/common/test_utils.py
@@ -417,9 +417,9 @@ class TestUtils(unittest.TestCase):
             o_count = 3
             b_used = 47
             return o_list, o_count, b_used
-        td = tempfile.mkdtemp()
         orig_gcd = utils.get_container_details
         utils.get_container_details = _mock_get_container_details
+        td = tempfile.mkdtemp()
         try:
             exp_md = {
                 utils.X_TYPE: (utils.CONTAINER, 0),
@@ -439,9 +439,9 @@ class TestUtils(unittest.TestCase):
             c_list = [ '123', 'abc' ]
             c_count = 2
             return c_list, c_count
-        td = tempfile.mkdtemp()
         orig_gad = utils.get_account_details
         utils.get_account_details = _mock_get_account_details
+        td = tempfile.mkdtemp()
         try:
             exp_md = {
                 utils.X_TYPE: (utils.ACCOUNT, 0),
@@ -752,12 +752,13 @@ class TestUtils(unittest.TestCase):
             utils.do_stat = orig_ds
 
     def test_get_container_details_from_fs(self):
-        td = tempfile.mkdtemp()
-        tf = tarfile.open("common/data/account_tree.tar.bz2", "r:bz2")
         orig_cwd = os.getcwd()
-        os.chdir(td)
-        tf.extractall()
+        td = tempfile.mkdtemp()
         try:
+            tf = tarfile.open("common/data/account_tree.tar.bz2", "r:bz2")
+            os.chdir(td)
+            tf.extractall()
+
             ad = utils._get_account_details_from_fs(td, None)
             assert ad.mtime == os.path.getmtime(td)
             assert ad.container_count == 3
@@ -775,12 +776,13 @@ class TestUtils(unittest.TestCase):
         assert cd.dir_list == []
 
     def test_get_account_details_from_fs(self):
-        td = tempfile.mkdtemp()
-        tf = tarfile.open("common/data/container_tree.tar.bz2", "r:bz2")
         orig_cwd = os.getcwd()
-        os.chdir(td)
-        tf.extractall()
+        td = tempfile.mkdtemp()
         try:
+            tf = tarfile.open("common/data/container_tree.tar.bz2", "r:bz2")
+            os.chdir(td)
+            tf.extractall()
+
             cd = utils._get_container_details_from_fs(td)
             assert cd.bytes_used == 30, repr(cd.bytes_used)
             assert cd.object_count == 8, repr(cd.object_count)
@@ -816,3 +818,32 @@ class TestUtils(unittest.TestCase):
         assert ad.mtime == os.path.getmtime(tf.name)
         assert ad.container_count == 0
         assert ad.container_list == []
+
+    def test_write_pickle(self):
+        td = tempfile.mkdtemp()
+        try:
+            fpp = os.path.join(td, 'pp')
+            utils.write_pickle('pickled peppers', fpp)
+            with open(fpp, "rb") as f:
+                contents = f.read()
+            s = pickle.loads(contents)
+            assert s == 'pickled peppers', repr(s)
+        finally:
+            shutil.rmtree(td)
+
+    def test_write_pickle_ignore_tmp(self):
+        tf = tempfile.NamedTemporaryFile()
+        td = tempfile.mkdtemp()
+        try:
+            fpp = os.path.join(td, 'pp')
+            # Also test an explicity pickle protocol
+            utils.write_pickle('pickled peppers', fpp, tmp=tf.name, pickle_protocol=2)
+            with open(fpp, "rb") as f:
+                contents = f.read()
+            s = pickle.loads(contents)
+            assert s == 'pickled peppers', repr(s)
+            with open(tf.name, "rb") as f:
+                contents = f.read()
+            assert contents == ''
+        finally:
+            shutil.rmtree(td)