tier/cli : printing a warning instead of skipping the node

        back-port of : http://review.gluster.org/#/c/14347/8

Problem: skipping the status of the nodes down creates confusion
to the user as one might see the status as completed for all nodes
and while performing detach commit, the operation will fail as the
node is down

Fix: Display a warning message

Note: When the last node is down (as per the peer list) then
warning message can't be displayed as the total number of peers
participating in the transaction is considered to be the total count.

>Change-Id: Ib7afbd1b26df3378e4d537db06f41f5c105ad86e
>BUG: 1324439
>Signed-off-by: hari gowtham <hgowtham@redhat.com>

Change-Id: Ie4296e932abaf163edc55b540b26dc6f5824ea85
BUG: 1328410
Signed-off-by: hari gowtham <hgowtham@redhat.com>
Reviewed-on: http://review.gluster.org/14458
Tested-by: hari gowtham <hari.gowtham005@gmail.com>
Smoke: Gluster Build System <jenkins@build.gluster.com>
NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org>
Reviewed-by: Atin Mukherjee <amukherj@redhat.com>
CentOS-regression: Gluster Build System <jenkins@build.gluster.com>

3 files changed, 40 insertions, 19 deletions

diff --git a/cli/src/cli-rpc-ops.c b/cli/src/cli-rpc-ops.c
index d836f8e6aee..8f6073fb1e8 100644
--- a/cli/src/cli-rpc-ops.c
+++ b/cli/src/cli-rpc-ops.c
@@ -1509,7 +1509,8 @@ out:
 }
 
 int
-gf_cli_print_rebalance_status (dict_t *dict, enum gf_task_types task_type)
+gf_cli_print_rebalance_status (dict_t *dict, enum gf_task_types task_type,
+                               gf_boolean_t is_tier)
 {
         int                ret          = -1;
         int                count        = 0;
@@ -1528,6 +1529,7 @@ gf_cli_print_rebalance_status (dict_t *dict, enum gf_task_types task_type)
         int                hrs          = 0;
         int                min          = 0;
         int                sec          = 0;
+        gf_boolean_t       down         = _gf_false;
 
         ret = dict_get_int32 (dict, "count", &count);
         if (ret) {
@@ -1562,6 +1564,7 @@ gf_cli_print_rebalance_status (dict_t *dict, enum gf_task_types task_type)
                         gf_log ("cli", GF_LOG_TRACE, "failed to get status");
                         gf_log ("cli", GF_LOG_ERROR, "node down and has failed"
                                 " to set dict");
+                        down = _gf_true;
                         continue;
                         /* skip this node if value not available*/
                 } else if (ret) {
@@ -1650,6 +1653,11 @@ gf_cli_print_rebalance_status (dict_t *dict, enum gf_task_types task_type)
                 }
                 GF_FREE(size_str);
         }
+        if (is_tier && down)
+                cli_out ("WARNING: glusterd might be down on one or more nodes."
+                         " Please check the nodes that are down using \'gluster"
+                         " peer status\' and start the glusterd on those nodes,"
+                         " else tier detach commit might fail!");
 out:
         return ret;
 }
@@ -1667,6 +1675,7 @@ gf_cli_print_tier_status (dict_t *dict, enum gf_task_types task_type)
         gf_defrag_status_t status_rcd   = GF_DEFRAG_STATUS_NOT_STARTED;
         char               *status_str  = NULL;
         char               *size_str    = NULL;
+        gf_boolean_t       down         = _gf_false;
 
         ret = dict_get_int32 (dict, "count", &count);
         if (ret) {
@@ -1695,6 +1704,7 @@ gf_cli_print_tier_status (dict_t *dict, enum gf_task_types task_type)
                                 "failed to get status", count, i);
                         gf_log ("cli", GF_LOG_ERROR, "node down and has failed"
                                 " to set dict");
+                        down = _gf_true;
                         continue;
                         /*skipping this node as value unavailable*/
                 } else if (ret) {
@@ -1733,8 +1743,11 @@ gf_cli_print_tier_status (dict_t *dict, enum gf_task_types task_type)
                 status_str = cli_vol_task_status_str[status_rcd];
                 cli_out ("%-20s %-20"PRIu64" %-20"PRIu64" %-20s",
                          node_name, promoted, demoted, status_str);
-
         }
+        if (down)
+                cli_out ("WARNING: glusterd might be down on one or more nodes."
+                         " Please check the nodes that are down using \'gluster"
+                         " peer status\' and start the glusterd on those nodes.");
 out:
         return ret;
 }
@@ -1893,9 +1906,14 @@ gf_cli_defrag_volume_cbk (struct rpc_req *req, struct iovec *iov,
 
         if (cmd == GF_DEFRAG_CMD_STATUS_TIER)
                 ret = gf_cli_print_tier_status (dict, GF_TASK_TYPE_REBALANCE);
+        else if (cmd == GF_DEFRAG_CMD_DETACH_STATUS)
+                ret = gf_cli_print_rebalance_status (dict,
+                                                     GF_TASK_TYPE_REBALANCE,
+                                                     _gf_true);
         else
                 ret = gf_cli_print_rebalance_status (dict,
-                                                     GF_TASK_TYPE_REBALANCE);
+                                                     GF_TASK_TYPE_REBALANCE,
+                                                     _gf_false);
 
         if (ret)
                 gf_log ("cli", GF_LOG_ERROR,
@@ -2491,7 +2509,8 @@ xml_output:
                 goto out;
         }
 
-        ret = gf_cli_print_rebalance_status (dict, GF_TASK_TYPE_REMOVE_BRICK);
+        ret = gf_cli_print_rebalance_status (dict, GF_TASK_TYPE_REMOVE_BRICK,
+                                             _gf_true);
         if (ret) {
                 gf_log ("cli", GF_LOG_ERROR, "Failed to print remove-brick "
                         "rebalance status");
@@ -2669,7 +2688,8 @@ xml_output:
                 goto out;
         }
 
-        ret = gf_cli_print_rebalance_status (dict, GF_TASK_TYPE_REMOVE_BRICK);
+        ret = gf_cli_print_rebalance_status (dict, GF_TASK_TYPE_REMOVE_BRICK,
+                                             _gf_false);
         if (ret) {
                 gf_log ("cli", GF_LOG_ERROR, "Failed to print remove-brick "
                         "rebalance status");
diff --git a/tests/basic/tier/new-tier-cmds.t b/tests/basic/tier/new-tier-cmds.t
index e4fef69d265..dbfac54938e 100644
--- a/tests/basic/tier/new-tier-cmds.t
+++ b/tests/basic/tier/new-tier-cmds.t
@@ -19,18 +19,6 @@ function create_dist_tier_vol () {
         TEST $CLI_1 volume attach-tier $V0 $H1:$B1/${V0}_h1 $H2:$B2/${V0}_h2 $H3:$B3/${V0}_h3
 }
 
-function tier_detach_commit () {
-	$CLI_1 volume tier $V0 detach commit | grep "success" | wc -l
-}
-
-function tier_detach_status_node_down () {
-        $CLI_1 volume tier $V0 detach status | wc -l
-}
-
-function tier_status_node_down () {
-	$CLI_1 volume tier $V0 status | wc -l
-}
-
 cleanup;
 
 #setup cluster and test volume
@@ -58,10 +46,10 @@ TEST $CLI_1 volume tier $V0 detach status
 TEST kill_node 2
 
 #check if we have the rest of the node available printed in the output of detach status
-EXPECT_WITHIN $PROCESS_UP_TIMEOUT "4" tier_detach_status_node_down
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "1" tier_detach_status_node_down
 
 #check if we have the rest of the node available printed in the output of tier status
-EXPECT_WITHIN $PROCESS_UP_TIMEOUT "5" tier_status_node_down
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "1" tier_status_node_down
 
 TEST $glusterd_2;
 
diff --git a/tests/tier.rc b/tests/tier.rc
index ee37e0704fa..69512c3fb4e 100644
--- a/tests/tier.rc
+++ b/tests/tier.rc
@@ -134,3 +134,16 @@ function rebalance_run_time () {
 
     echo $total;
 }
+
+function tier_detach_commit () {
+	$CLI_1 volume tier $V0 detach commit | grep "success" | wc -l
+}
+
+function tier_detach_status_node_down () {
+        $CLI_1 volume tier $V0 detach status | grep "WARNING" | wc -l
+}
+
+function tier_status_node_down () {
+	$CLI_1 volume tier $V0 status | grep "WARNING" | wc -l
+}
+