Rebase to stable/kilo

This change ports most of swiftonfile object server fixes and changes
into gluster-swift. Storage policy as a feature is not usable here
(it doesn't make sense).

The hacky way of creating zero byte tracker objects for object
expiration has not been ported to this release due to scalability
issues and the need to have a separate volume.

Change-Id: I17ba27dacea9ac000bdb8934700996e4d17f4251
Signed-off-by: Prashanth Pai <ppai@redhat.com>
Reviewed-on: http://review.gluster.org/13269
Reviewed-by: Thiago da Silva <thiago@redhat.com>
Tested-by: Thiago da Silva <thiago@redhat.com>

4 files changed, 105 insertions, 29 deletions

diff --git a/gluster/swift/common/DiskDir.py b/gluster/swift/common/DiskDir.py
index 6112709..e8dba35 100644
--- a/gluster/swift/common/DiskDir.py
+++ b/gluster/swift/common/DiskDir.py
@@ -255,7 +255,7 @@ class DiskDir(DiskCommon):
     :param uid: user ID container object should assume
     :param gid: group ID container object should assume
 
-    Usage pattern from container/server.py (Havana, 1.8.0+):
+    Usage pattern from container/server.py (Kilo, 2.3.0):
         DELETE:
             if auto-create and obj and not .db_file:
                 # Creates container
@@ -287,36 +287,43 @@ class DiskDir(DiskCommon):
                     return 404
                 .put_object()
             else:
-                if not .db_file:
-                    # Creates container
-                    .initialize()
-                else:
-                    # Update container timestamp
-                    .is_deleted()
+                _update_or_create():
+                    if not .db_file:
+                        # Creates container
+                        .initialize()
+                    recreated = .is_deleted():
+                    if recreated:
+                        .set_storage_policy_index()
+                    .storage_policy_index
                     .update_put_timestamp()
                     if .is_deleted()
                         return conflict
-                if metadata:
+                    if recreated:
+                        .update_status_changed_at()
+
+                if 'X-Container-Sync-To' in metadata:
                     if .metadata
                         .set_x_container_sync_points()
                     .update_metadata()
                 account_update():
                     .get_info()
         HEAD:
-            .pending_timeout
-            .stale_reads_ok
-            if .is_deleted():
-                return 404
-            .get_info()
+            info, is_deleted = .get_info_is_deleted()
+            .get_info_is_deleted():
+                if not .db_file:
+                    return {}, True
+                info = .get_info()
+                return info, ._is_deleted_info()
             .metadata
         GET:
-            .pending_timeout
-            .stale_reads_ok
-            if .is_deleted():
-                return 404
-            .get_info()
-            .metadata
+            info, is_deleted = .get_info_is_deleted()
+            .get_info_is_deleted():
+                if not .db_file:
+                    return {}, True
+                info = .get_info()
+                return info, ._is_deleted_info()
             .list_objects_iter()
+            .metadata
         POST:
             if .is_deleted():
                 return 404
@@ -346,8 +353,22 @@ class DiskDir(DiskCommon):
                 create_container_metadata(self.datadir)
                 self.metadata = _read_metadata(self.datadir)
 
+    def update_status_changed_at(self, timestamp):
+        return
+
+    @property
+    def storage_policy_index(self):
+        if not hasattr(self, '_storage_policy_index'):
+            self._storage_policy_index = \
+                self.get_info()['storage_policy_index']
+        return self._storage_policy_index
+
+    def set_storage_policy_index(self, policy_index, timestamp=None):
+        self._storage_policy_index = policy_index
+
     def list_objects_iter(self, limit, marker, end_marker,
-                          prefix, delimiter, path=None):
+                          prefix, delimiter, path=None,
+                          storage_policy_index=0):
         """
         Returns tuple of name, created_at, size, content_type, etag.
         """
@@ -451,6 +472,12 @@ class DiskDir(DiskCommon):
 
         return objects
 
+    def get_info_is_deleted(self):
+        if not do_exists(self.datadir):
+            return {}, True
+        info = self.get_info()
+        return info, False
+
     def get_info(self):
         """
         Get global data for the container.
@@ -477,7 +504,10 @@ class DiskDir(DiskCommon):
                     'x_container_sync_point1', -1),
                 'x_container_sync_point2': self.metadata.get(
                     'x_container_sync_point2', -1),
+                'storage_policy_index': self.metadata.get(
+                    'storage_policy_index', 0)
                 }
+        self._storage_policy_index = data['storage_policy_index']
         return data
 
     def put_object(self, name, timestamp, size, content_type, etag, deleted=0):
@@ -540,13 +570,14 @@ class DiskDir(DiskCommon):
 
 class DiskAccount(DiskCommon):
     """
-    Usage pattern from account/server.py (Havana, 1.8.0+):
+    Usage pattern from account/server.py (Kilo, 2.3.0):
         DELETE:
             .is_deleted()
+            .is_status_deleted()
             .delete_db()
+            .is_status_deleted()
         PUT:
             container:
-                .pending_timeout
                 .db_file
                 .initialize()
                 .is_deleted()
@@ -555,25 +586,27 @@ class DiskAccount(DiskCommon):
                 .db_file
                 .initialize()
                 .is_status_deleted()
+                .is_status_deleted()
                 .is_deleted()
                 .update_put_timestamp()
-                .is_deleted() ???
+                .is_deleted()
                 .update_metadata()
         HEAD:
-            .pending_timeout
-            .stale_reads_ok
             .is_deleted()
+            .is_status_deleted()
             .get_info()
+            .get_policy_stats()
             .metadata
         GET:
-            .pending_timeout
-            .stale_reads_ok
             .is_deleted()
+            .is_status_deleted()
             .get_info()
+            .get_policy_stats()
             .metadata
             .list_containers_iter()
         POST:
             .is_deleted()
+            .is_status_deleted()
             .update_metadata()
     """
 
@@ -748,3 +781,6 @@ class DiskAccount(DiskCommon):
                 'bytes_used': self.metadata.get(X_BYTES_USED, (0, 0))[0],
                 'hash': '', 'id': ''}
         return data
+
+    def get_policy_stats(self, do_migrations=False):
+        return {}
diff --git a/gluster/swift/common/Glusterfs.py b/gluster/swift/common/Glusterfs.py
index 6a2fdb2..910f752 100644
--- a/gluster/swift/common/Glusterfs.py
+++ b/gluster/swift/common/Glusterfs.py
@@ -148,7 +148,7 @@ def _get_unique_id():
                 # own the lock.
                 continue
             raise
-        except:
+        except Exception:
             os.close(fd)
             raise
         else:
diff --git a/gluster/swift/common/middleware/gswauth/swauth/middleware.py b/gluster/swift/common/middleware/gswauth/swauth/middleware.py
index cdcc638..3cd9cf7 100644
--- a/gluster/swift/common/middleware/gswauth/swauth/middleware.py
+++ b/gluster/swift/common/middleware/gswauth/swauth/middleware.py
@@ -28,6 +28,7 @@ import base64
 
 from eventlet.timeout import Timeout
 from eventlet import TimeoutError
+from swift import gettext_ as _
 from swift.common.swob import HTTPAccepted, HTTPBadRequest, HTTPConflict, \
     HTTPCreated, HTTPForbidden, HTTPMethodNotAllowed, HTTPMovedPermanently, \
     HTTPNoContent, HTTPNotFound, HTTPUnauthorized, \
diff --git a/gluster/swift/common/utils.py b/gluster/swift/common/utils.py
index b6a5a09..e6f4bcc 100644
--- a/gluster/swift/common/utils.py
+++ b/gluster/swift/common/utils.py
@@ -17,6 +17,7 @@ import os
 import stat
 import json
 import errno
+import random
 import logging
 from hashlib import md5
 from eventlet import sleep
@@ -29,7 +30,7 @@ from swift.common.db import utf8encodekeys
 from gluster.swift.common.fs_utils import do_getctime, do_getmtime, do_stat, \
     do_listdir, do_walk, do_rmdir, do_log_rl, get_filename_from_fd, do_open, \
     do_isdir, do_getsize, do_getxattr, do_setxattr, do_removexattr, do_read, \
-    do_close, do_dup, do_lseek, do_fstat
+    do_close, do_dup, do_lseek, do_fstat, do_fsync, do_rename
 from gluster.swift.common import Glusterfs
 
 X_CONTENT_TYPE = 'Content-Type'
@@ -607,3 +608,41 @@ def rmobjdir(dir_path):
         raise
     else:
         return True
+
+
+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)
+    # Create destination directory
+    try:
+        os.makedirs(dirname)
+    except OSError as err:
+        if err.errno != errno.EEXIST:
+            raise
+    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)
+        # TODO: This flush() method call turns into a flush() system call
+        # We'll need to wrap this as well, but we would do this by writing
+        # a context manager for our own open() method which returns an object
+        # in fo which makes the gluster API call.
+        fo.flush()
+        do_fsync(fo)
+    do_rename(tmppath, dest)