Tier/glusterd: Resetting the tier status value to not started

        back-port of : http://review.gluster.org/#/c/14106/

Problem: during a volume restart or tier start force, the
value of tier status is set as started irrespective of the result.

Fix: The appropriate value of status is set during the restart of
rebalance function.

>Change-Id: I6164f0add48542a57dee059e80fa0f9bb036dbef
>BUG: 1315666
>Signed-off-by: hari <hgowtham@redhat.com>

Change-Id: Ie4345bd7ce1d458574e36b70fe8994b3d758396a
BUG: 1316808
Signed-off-by: hari <hgowtham@redhat.com>
Reviewed-on: http://review.gluster.org/14229
Smoke: Gluster Build System <jenkins@build.gluster.com>
Tested-by: hari gowtham <hari.gowtham005@gmail.com>
NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org>
CentOS-regression: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Atin Mukherjee <amukherj@redhat.com>

2 files changed, 67 insertions, 22 deletions

diff --git a/tests/basic/tier/tierd_check.t b/tests/basic/tier/tierd_check.t
index 2b8ccbbbd96..1f88ea0b72e 100644
--- a/tests/basic/tier/tierd_check.t
+++ b/tests/basic/tier/tierd_check.t
@@ -3,20 +3,24 @@
 . $(dirname $0)/../../include.rc
 . $(dirname $0)/../../volume.rc
 . $(dirname $0)/../../tier.rc
+. $(dirname $0)/../../cluster.rc
 
 
 # Creates a tiered volume with pure distribute hot and cold tiers
 # Both hot and cold tiers will have an equal number of bricks.
 
+function check_peers {
+    $CLI_1 peer status | grep 'Peer in Cluster (Connected)' | wc -l
+}
+
 function create_dist_tier_vol () {
-        mkdir $B0/cold
-        mkdir $B0/hot
-        TEST $CLI volume create $V0 $H0:$B0/cold/${V0}{1..3}
-        TEST $CLI volume set $V0 performance.quick-read off
-        TEST $CLI volume set $V0 performance.io-cache off
-        TEST $CLI volume start $V0
-        TEST $CLI volume attach-tier $V0 $H0:$B0/hot/${V0}{1..2}
-        TEST $CLI volume set $V0 cluster.tier-mode test
+        TEST $CLI_1 volume create $V0 $H1:$B1/${V0} $H2:$B2/${V0}
+        TEST $CLI_1 volume start $V0
+        TEST $CLI_1 volume attach-tier $V0 $H1:$B1/${V0}_h1 $H2:$B2/${V0}_h2
+}
+
+function tier_status () {
+	$CLI_1 volume tier $V0 status | grep progress | wc -l
 }
 
 function tier_deamon_kill () {
@@ -26,38 +30,74 @@ echo "$?"
 
 cleanup;
 
-#Basic checks
-TEST glusterd
-TEST pidof glusterd
-TEST $CLI volume status
+#setup cluster and test volume
+TEST launch_cluster 3; # start 3-node virtual cluster
+TEST $CLI_1 peer probe $H2; # peer probe server 2 from server 1 cli
+TEST $CLI_1 peer probe $H3; # peer probe server 3 from server 1 cli
 
+EXPECT_WITHIN $PROBE_TIMEOUT 2 check_peers;
 
 #Create and start a tiered volume
 create_dist_tier_vol
 
 EXPECT_WITHIN $PROCESS_UP_TIMEOUT 0 tier_daemon_check
 
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "2" tier_status
+
 EXPECT_WITHIN $PROCESS_UP_TIMEOUT 0 tier_deamon_kill
 
-TEST $CLI volume tier $V0 start
+TEST $CLI_1 volume tier $V0 start
 
 EXPECT_WITHIN $PROCESS_UP_TIMEOUT "0" tier_daemon_check
 
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "2" tier_status
+
 EXPECT_WITHIN $PROCESS_UP_TIMEOUT "0" tier_deamon_kill
 
-TEST $CLI volume tier $V0 start force
+TEST $CLI_3 volume tier $V0 start force
 
 EXPECT_WITHIN $PROCESS_UP_TIMEOUT "0" tier_daemon_check
 
-TEST $CLI volume tier $V0 start force
+#The pattern progress should occur twice only.
+#it shouldn't come up on the third node without tierd even
+#after the tier start force is issued on the node without
+#tierd
+
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "2" tier_status
+
+#kill the node on which tier is not supposed to run
+TEST kill_node 3
+
+#bring the node back, it should not have tierd running on it
+TEST $glusterd_3;
+
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "2" tier_status
+
+#after volume restart, check for tierd
+
+TEST $CLI_3 volume stop $V0
+
+TEST $CLI_3 volume start $V0
+
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "2" tier_status
+
+#check for detach start and stop
+
+TEST $CLI_3 volume tier $V0 detach start
+
+TEST $CLI_3 volume tier $V0 detach stop
+
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "2" tier_status
+
+TEST $CLI_1 volume tier $V0 start force
 
 EXPECT_WITHIN $PROCESS_UP_TIMEOUT "0" tier_daemon_check
 
-# To test fordetach start fail while the brick is down
+# To test for detach start fail while the brick is down
 
-TEST pkill -f "$B0/hot/$V0"
+TEST pkill -f "$B1/$V0"
 
-TEST ! $CLI volume tier $V0 detach start
+TEST ! $CLI_1 volume tier $V0 detach start
 
 cleanup
 #G_TESTDEF_TEST_STATUS_NETBSD7=KNOWN_ISSUE,BUG=000000
diff --git a/xlators/mgmt/glusterd/src/glusterd-utils.c b/xlators/mgmt/glusterd/src/glusterd-utils.c
index 43a4898536b..99b01e32915 100644
--- a/xlators/mgmt/glusterd/src/glusterd-utils.c
+++ b/xlators/mgmt/glusterd/src/glusterd-utils.c
@@ -7225,6 +7225,8 @@ glusterd_volume_defrag_restart (glusterd_volinfo_t *volinfo, char *op_errstr,
         case GF_DEFRAG_STATUS_NOT_STARTED:
                 ret = glusterd_handle_defrag_start (volinfo, op_errstr, len,
                                 cmd, cbk, volinfo->rebal.op);
+                if (ret)
+                        volinfo->rebal.defrag_status = GF_DEFRAG_STATUS_FAILED;
                 break;
         default:
                 gf_msg (this->name, GF_LOG_ERROR, 0,
@@ -7236,6 +7238,7 @@ glusterd_volume_defrag_restart (glusterd_volinfo_t *volinfo, char *op_errstr,
         }
 out:
         return ret;
+
 }
 
 void
@@ -7297,9 +7300,6 @@ glusterd_restart_rebalance_for_volume (glusterd_volinfo_t *volinfo)
         int             ret = -1;
         char          op_errstr[PATH_MAX];
 
-        if (!volinfo->rebal.defrag_cmd)
-                return -1;
-
         if (!gd_should_i_start_rebalance (volinfo)) {
 
                 /* Store the rebalance-id and rebalance command even if
@@ -7310,11 +7310,17 @@ glusterd_restart_rebalance_for_volume (glusterd_volinfo_t *volinfo)
                  * Storing this is needed for having 'volume status'
                  * work correctly.
                  */
+                volinfo->rebal.defrag_status = GF_DEFRAG_STATUS_NOT_STARTED;
                 if (volinfo->type == GF_CLUSTER_TYPE_TIER)
                         glusterd_store_perform_node_state_store (volinfo);
 
                 return 0;
         }
+        if (!volinfo->rebal.defrag_cmd) {
+                volinfo->rebal.defrag_status = GF_DEFRAG_STATUS_FAILED;
+                return -1;
+        }
+
         ret = glusterd_volume_defrag_restart (volinfo, op_errstr, PATH_MAX,
                                 volinfo->rebal.defrag_cmd,
                                 volinfo->rebal.op == GD_OP_REMOVE_BRICK ?
@@ -7329,7 +7335,6 @@ glusterd_restart_rebalance_for_volume (glusterd_volinfo_t *volinfo)
                         volinfo->decommission_in_progress = 1;
                 }
         }
-
         return ret;
 }
 int