georep: add reset-sync-time option for session delete

Set the stime xattr at all the brick roots to (0,0) if the argument
reset-sync-time has been provided on the command-line.
To avoid testing against directory specific stime, the remote
stime is assumed to be minus_infinity, if the root directory
stime is set to (0,0), before the directory scan begins.
This triggers a full volume resync to slave in the case of a
geo-rep session recreation with the same master-slave volume
pair.

Command synopsis:
gluster volume geo-replication <MASTERVOL> <SLAVE>::<SLAVEVOL> delete \
    [reset-sync-time]

Update gluster cli man page to include new sub-command reset-sync-time.

Change-Id: Ie4ce03b9425ed9bb81eda8681058c0fc6f990948
BUG: 1311926
Signed-off-by: Milind Changire <mchangir@redhat.com>
Reviewed-on: http://review.gluster.org/14051
Reviewed-by: Kotresh HR <khiremat@redhat.com>
Smoke: Gluster Build System <jenkins@build.gluster.org>
NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org>
CentOS-regression: Gluster Build System <jenkins@build.gluster.org>
Reviewed-by: Aravinda VK <avishwan@redhat.com>

7 files changed, 67 insertions, 6 deletions

diff --git a/cli/src/cli-cmd-parser.c b/cli/src/cli-cmd-parser.c
index ae74d441c33..a570c343133 100644
--- a/cli/src/cli-cmd-parser.c
+++ b/cli/src/cli-cmd-parser.c
@@ -2574,7 +2574,7 @@ cli_cmd_gsync_set_parse (const char **words, int wordcount, dict_t **options)
          * volume geo-replication [$m [$s]] status [detail]
          * volume geo-replication [$m] $s config [[!]$opt [$val]]
          * volume geo-replication $m $s start|stop [force]
-         * volume geo-replication $m $s delete
+         * volume geo-replication $m $s delete [reset-sync-time]
          * volume geo-replication $m $s pause [force]
          * volume geo-replication $m $s resume [force]
          */
@@ -2702,6 +2702,22 @@ cli_cmd_gsync_set_parse (const char **words, int wordcount, dict_t **options)
                 cmdi++;
         }
 
+        if (type == GF_GSYNC_OPTION_TYPE_DELETE &&
+            !strcmp ((char *)words[wordcount-1], "reset-sync-time")) {
+                if (strcmp ((char *)words[wordcount-2], "delete")) {
+                        ret = -1;
+                        goto out;
+                }
+                if (!slavei || !masteri) {
+                        ret = -1;
+                        goto out;
+                }
+                ret = dict_set_uint32 (dict, "reset-sync-time", _gf_true);
+                if (ret)
+                        goto out;
+                cmdi++;
+        }
+
         if (type != GF_GSYNC_OPTION_TYPE_CONFIG &&
             (cmdi < wordcount - 1 || glob))
                 goto out;
diff --git a/cli/src/cli-cmd-volume.c b/cli/src/cli-cmd-volume.c
index b09a6ccbe3d..2ae4aa2dbf4 100644
--- a/cli/src/cli-cmd-volume.c
+++ b/cli/src/cli-cmd-volume.c
@@ -2636,7 +2636,7 @@ struct cli_cmd volume_cmds[] = {
 
 #if (SYNCDAEMON_COMPILE)
         {"volume "GEOREP" [<VOLNAME>] [<SLAVE-URL>] {create [[ssh-port n] [[no-verify]|[push-pem]]] [force]"
-         "|start [force]|stop [force]|pause [force]|resume [force]|config|status [detail]|delete} [options...]",
+         "|start [force]|stop [force]|pause [force]|resume [force]|config|status [detail]|delete [reset-sync-time]} [options...]",
          cli_cmd_volume_gsync_set_cbk,
          "Geo-sync operations",
          cli_cmd_check_gsync_exists_cbk},
diff --git a/doc/gluster.8 b/doc/gluster.8
index e726010ee5d..1c6cf09fdd5 100644
--- a/doc/gluster.8
+++ b/doc/gluster.8
@@ -144,8 +144,9 @@ Query status of the geo-replication session from <MASTER_VOL> to <SLAVE_HOST> ho
 \fB\ volume geo-replication <MASTER_VOL> <SLAVE_HOST>::<SLAVE_VOL> {pause|resume} [force] \fR
 Pause/resume the geo-replication session from <MASTER_VOL> to <SLAVE_HOST> host machine having <SLAVE_VOL>.
 .TP
-\fB\ volume geo-replication <MASTER_VOL> <SLAVE_HOST>::<SLAVE_VOL> delete \fR
+\fB\ volume geo-replication <MASTER_VOL> <SLAVE_HOST>::<SLAVE_VOL> delete [reset-sync-time]\fR
 Delete the geo-replication session from <MASTER_VOL> to <SLAVE_HOST> host machine having <SLAVE_VOL>.
+Optionally you can also reset the sync time in case you need to resync the entire volume on session recreate.
 .TP
 \fB\ volume geo-replication <MASTER_VOL> <SLAVE_HOST>::<SLAVE_VOL> config [[!]<options> [<value>]] \fR
 View (when no option provided) or set configuration for this geo-replication session.
diff --git a/geo-replication/syncdaemon/gsyncd.py b/geo-replication/syncdaemon/gsyncd.py
index da5de0f40f8..f41756a0b93 100644
--- a/geo-replication/syncdaemon/gsyncd.py
+++ b/geo-replication/syncdaemon/gsyncd.py
@@ -27,7 +27,7 @@ from ipaddr import IPAddress, IPNetwork
 
 from gconf import gconf
 from syncdutils import FreeObject, norm, grabpidfile, finalize
-from syncdutils import log_raise_exception, privileged
+from syncdutils import log_raise_exception, privileged, boolify
 from syncdutils import GsyncdError, select, set_term_handler
 from configinterface import GConffile, upgrade_config_file
 import resource
@@ -37,6 +37,8 @@ from subprocess import PIPE
 import subprocess
 from changelogagent import agent, Changelog
 from gsyncdstatus import set_monitor_status, GeorepStatus
+from libcxattr import Xattr
+import struct
 
 ParseError = XET.ParseError if hasattr(XET, 'ParseError') else SyntaxError
 
@@ -253,9 +255,9 @@ def main_i():
                   action='callback', callback=store_abs)
     op.add_option('-l', '--log-file', metavar='LOGF', type=str,
                   action='callback', callback=store_abs)
-    op.add_option('--iprefix',  metavar='LOGD',  type=str,
+    op.add_option('--iprefix', metavar='LOGD', type=str,
                   action='callback', callback=store_abs)
-    op.add_option('--changelog-log-file',  metavar='LOGF',  type=str,
+    op.add_option('--changelog-log-file', metavar='LOGF', type=str,
                   action='callback', callback=store_abs)
     op.add_option('--log-file-mbr', metavar='LOGF', type=str,
                   action='callback', callback=store_abs)
@@ -355,6 +357,9 @@ def main_i():
                   action='callback', callback=store_local)
     op.add_option('--delete', dest='delete', action='callback',
                   callback=store_local_curry(True))
+    op.add_option('--path-list', dest='path_list', action='callback',
+                  type=str, callback=store_local)
+    op.add_option('--reset-sync-time', default=False, action='store_true')
     op.add_option('--status-get', dest='status_get', action='callback',
                   callback=store_local_curry(True))
     op.add_option('--debug', dest="go_daemon", action='callback',
@@ -572,6 +577,10 @@ def main_i():
     delete = rconf.get('delete')
     if delete:
         logging.info('geo-replication delete')
+        # remove the stime xattr from all the brick paths so that
+        # a re-create of a session will start sync all over again
+        stime_xattr_name = getattr(gconf, 'master.stime_xattr_name', None)
+
         # Delete pid file, status file, socket file
         cleanup_paths = []
         if getattr(gconf, 'pid_file', None):
@@ -604,6 +613,20 @@ def main_i():
             # To delete temp files
             for f in glob.glob(path + "*"):
                 _unlink(f)
+
+        reset_sync_time = boolify(gconf.reset_sync_time)
+        if reset_sync_time and stime_xattr_name:
+            path_list = rconf.get('path_list')
+            paths = []
+            for p in path_list.split('--path='):
+                stripped_path = p.strip()
+                if stripped_path != "":
+                    # set stime to (0,0) to trigger full volume content resync
+                    # to slave on session recreation
+                    # look at master.py::Xcrawl   hint: zero_zero
+                    Xattr.lsetxattr(stripped_path, stime_xattr_name,
+                                    struct.pack("!II", 0, 0))
+
         return
 
     if restricted and gconf.allow_network:
diff --git a/geo-replication/syncdaemon/master.py b/geo-replication/syncdaemon/master.py
index be68a7fd7b4..80c4d9d8b95 100644
--- a/geo-replication/syncdaemon/master.py
+++ b/geo-replication/syncdaemon/master.py
@@ -1377,6 +1377,9 @@ class GMasterXsyncMixin(GMasterChangelogMixin):
                              "correct xtime for %s (%d)" % (path, xtr))
             xtr = self.minus_infinity
         xtr = max(xtr, xtr_root)
+        zero_zero = (0, 0)
+        if xtr_root == zero_zero:
+            xtr = self.minus_infinity
         if not self.need_sync(path, xtl, xtr):
             if path == '.':
                 self.sync_done([(path, xtl)], True)
diff --git a/geo-replication/syncdaemon/monitor.py b/geo-replication/syncdaemon/monitor.py
index 050218b6d1b..a26de0c9cf5 100644
--- a/geo-replication/syncdaemon/monitor.py
+++ b/geo-replication/syncdaemon/monitor.py
@@ -165,6 +165,7 @@ class Volinfo(object):
         else:
             return 0
 
+
 class Monitor(object):
 
     """class which spawns and manages gsyncd workers"""
@@ -428,6 +429,15 @@ def distribute(*resources):
         suuid = svol.uuid
         slave_host = slave.remote_addr.split('@')[-1]
         slave_vol = si.volume
+
+        # save this xattr for the session delete command
+        old_stime_xattr_name = getattr(gconf, "master.stime_xattr_name", None)
+        new_stime_xattr_name = "trusted.glusterfs." + mvol.uuid + "." + \
+            svol.uuid + ".stime"
+        if not old_stime_xattr_name or \
+           old_stime_xattr_name != new_stime_xattr_name:
+            gconf.configinterface.set("master.stime_xattr_name",
+                                      new_stime_xattr_name)
     else:
         raise GsyncdError("unknown slave type " + slave.url)
     logging.info('slave bricks: ' + repr(sbricks))
diff --git a/xlators/mgmt/glusterd/src/glusterd-geo-rep.c b/xlators/mgmt/glusterd/src/glusterd-geo-rep.c
index 57b50934489..4a0ef095efd 100644
--- a/xlators/mgmt/glusterd/src/glusterd-geo-rep.c
+++ b/xlators/mgmt/glusterd/src/glusterd-geo-rep.c
@@ -5005,6 +5005,7 @@ glusterd_gsync_delete (glusterd_volinfo_t *volinfo, char *slave,
         char             geo_rep_dir[PATH_MAX] = "";
         char            *conf_path = NULL;
         xlator_t *this = NULL;
+        uint32_t        reset_sync_time = _gf_false;
 
         this = THIS;
         GF_ASSERT (this);
@@ -5041,6 +5042,13 @@ glusterd_gsync_delete (glusterd_volinfo_t *volinfo, char *slave,
         runner_argprintf (&runner, "%s", conf_path);
         runner_argprintf (&runner, "--iprefix=%s", DATADIR);
 
+        runner_argprintf (&runner, "--path-list=%s", path_list);
+
+        ret = dict_get_uint32 (dict, "reset-sync-time", &reset_sync_time);
+        if (!ret && reset_sync_time) {
+                runner_add_args  (&runner, "--reset-sync-time", NULL);
+        }
+
         if (volinfo) {
                 master = volinfo->volname;
                 runner_argprintf (&runner, ":%s", master);