geo-rep: fd close and fcntl issue

When any of the open fd of a file is closed
on which fcntl lock is taken even though another
fd of the same file is open on which lock is taken,
all fcntl locks will be released. This causes both
replica workers to be ACTIVE sometimes. This patche
fixes that issue.

Change-Id: I1e203ab0e29442275338276deb56d09e5679329c
BUG: 1285488
Original-Author: Aravinda VK <avishwan@redhat.com>
Signed-off-by: Kotresh HR <khiremat@redhat.com>
Reviewed-on: http://review.gluster.org/12752
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Milind Changire <mchangir@redhat.com>
Tested-by: NetBSD Build System <jenkins@build.gluster.org>
Reviewed-by: Aravinda VK <avishwan@redhat.com>
Reviewed-by: Venky Shankar <vshankar@redhat.com>

1 files changed, 19 insertions, 15 deletions

diff --git a/geo-replication/syncdaemon/master.py b/geo-replication/syncdaemon/master.py
index 55226bdf94f..eadf51d3e5b 100644
--- a/geo-replication/syncdaemon/master.py
+++ b/geo-replication/syncdaemon/master.py
@@ -443,6 +443,24 @@ class GMasterCommon(object):
     def mgmt_lock(self):
 
         """Take management volume lock """
+        if gconf.mgmt_lock_fd:
+            try:
+                fcntl.lockf(gconf.mgmt_lock_fd, fcntl.LOCK_EX | fcntl.LOCK_NB)
+                if not gconf.active_earlier:
+                    gconf.active_earlier = True
+                    logging.info("Got lock : %s : Becoming ACTIVE"
+                                 % gconf.local_path)
+                return True
+            except:
+                ex = sys.exc_info()[1]
+                if isinstance(ex, IOError) and ex.errno in (EACCES, EAGAIN):
+                    if not gconf.passive_earlier:
+                        gconf.passive_earlier = True
+                        logging.info("Didn't get lock : %s : Becoming PASSIVE"
+                                     % gconf.local_path)
+                    return False
+                raise
+
         fd = None
         bname = str(self.uuid) + "_" + str(gconf.slave_id) + "_subvol_" \
             + str(gconf.subvol_num) + ".lock"
@@ -468,31 +486,17 @@ class GMasterCommon(object):
                 raise
         try:
             fcntl.lockf(fd, fcntl.LOCK_EX | fcntl.LOCK_NB)
-            # Close the previously acquired lock so that
-            # fd will not leak. Reset fd to None
-            if gconf.mgmt_lock_fd:
-                os.close(gconf.mgmt_lock_fd)
-                gconf.mgmt_lock_fd = None
-
             # Save latest FD for future use
             gconf.mgmt_lock_fd = fd
         except:
             ex = sys.exc_info()[1]
-            if fd:
-                os.close(fd)
-
-            # When previously Active becomes Passive, Close the
-            # fd of previously acquired lock
-            if gconf.mgmt_lock_fd:
-                os.close(gconf.mgmt_lock_fd)
-                gconf.mgmt_lock_fd = None
-
             if isinstance(ex, IOError) and ex.errno in (EACCES, EAGAIN):
                 # cannot grab, it's taken
                 if not gconf.passive_earlier:
                     gconf.passive_earlier = True
                     logging.info("Didn't get lock : %s : Becoming PASSIVE"
                                  % gconf.local_path)
+                gconf.mgmt_lock_fd = fd
                 return False
             raise