glusterd: Added volume-id to 'op' dictionary

Volume-id passed in op dictionary would help detect possible split brains
among peers in a cluster. The idea is to check if the volume's id and
the vol-id that was passed are equal.
ie, same volume name, but different volume id indicate that glusterd
'metadata' of one of the participating peers is stale or there is
a split brain.

This is over and above the existing checksum based validation of peer
supplied cluster 'metadata' (ie, volume info file).

Change-Id: I1049ef249e417e540ccb4243e450f92fcd0f46f9
BUG: 797734
Signed-off-by: Krishnan Parthasarathi <kp@gluster.com>
Reviewed-on: http://review.gluster.com/3083
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Jeff Darcy <jdarcy@redhat.com>
Reviewed-by: Vijay Bellur <vijay@gluster.com>

1 files changed, 53 insertions, 0 deletions

diff --git a/xlators/mgmt/glusterd/src/glusterd-op-sm.c b/xlators/mgmt/glusterd/src/glusterd-op-sm.c
index a577e161c17..1825065a4b4 100644
--- a/xlators/mgmt/glusterd/src/glusterd-op-sm.c
+++ b/xlators/mgmt/glusterd/src/glusterd-op-sm.c
@@ -504,6 +504,7 @@ glusterd_op_stage_reset_volume (dict_t *dict, char **op_errstr)
         char                                    msg[2048]      = {0};
         char                                    *key = NULL;
         char                                    *key_fixed = NULL;
+        glusterd_volinfo_t                      *volinfo       = NULL;
 
         ret = dict_get_str (dict, "volname", &volname);
 
@@ -522,6 +523,13 @@ glusterd_op_stage_reset_volume (dict_t *dict, char **op_errstr)
                 ret = -1;
                 goto out;
         }
+        ret = glusterd_volinfo_find (volname, &volinfo);
+        if (ret)
+                goto out;
+
+        ret = glusterd_validate_volume_id (dict, volinfo);
+        if (ret)
+                goto out;
 
         ret = dict_get_str (dict, "key", &key);
         if (ret) {
@@ -568,6 +576,7 @@ glusterd_op_stage_sync_volume (dict_t *dict, char **op_errstr)
         gf_boolean_t                            exists = _gf_false;
         glusterd_peerinfo_t                     *peerinfo = NULL;
         char                                    msg[2048] = {0,};
+        glusterd_volinfo_t                      *volinfo  = NULL;
 
         ret = dict_get_str (dict, "hostname", &hostname);
         if (ret) {
@@ -607,6 +616,13 @@ glusterd_op_stage_sync_volume (dict_t *dict, char **op_errstr)
                                 ret = -1;
                                 goto out;
                         }
+                        ret = glusterd_volinfo_find (volname, &volinfo);
+                        if (ret)
+                                goto out;
+
+                        ret = glusterd_validate_volume_id (dict, volinfo);
+                        if (ret)
+                                goto out;
                 } else {
                         ret = 0;
                 }
@@ -664,6 +680,10 @@ glusterd_op_stage_status_volume (dict_t *dict, char **op_errstr)
                 goto out;
         }
 
+        ret = glusterd_validate_volume_id (dict, volinfo);
+        if (ret)
+                goto out;
+
         ret = glusterd_is_volume_started (volinfo);
         if (!ret) {
                 snprintf (msg, sizeof (msg), "Volume %s is not started",
@@ -799,6 +819,10 @@ glusterd_op_stage_stats_volume (dict_t *dict, char **op_errstr)
                 goto out;
         }
 
+        ret = glusterd_validate_volume_id (dict, volinfo);
+        if (ret)
+                goto out;
+
         ret = dict_get_int32 (dict, "op", &stats_op);
         if (ret) {
                 snprintf (msg, sizeof (msg), "Volume profile op get failed");
@@ -1771,6 +1795,9 @@ glusterd_op_build_payload (dict_t **req)
         void                    *ctx = NULL;
         dict_t                  *req_dict = NULL;
         glusterd_op_t           op = GD_OP_NONE;
+        char                    *volname = NULL;
+        char                    *volid = NULL;
+        glusterd_volinfo_t      *volinfo = NULL;
 
         GF_ASSERT (req);
 
@@ -1820,6 +1847,32 @@ glusterd_op_build_payload (dict_t **req)
                 case GD_OP_DEFRAG_BRICK_VOLUME:
                         {
                                 dict_t  *dict = ctx;
+                                ret = dict_get_str (dict, "volname", &volname);
+                                if (ret) {
+                                        gf_log (THIS->name, GF_LOG_CRITICAL,
+                                                "volname is not present in "
+                                                "operation ctx");
+                                        goto out;
+                                }
+                                ret = glusterd_volinfo_find (volname, &volinfo);
+                                if (ret) {
+                                        gf_log (THIS->name, GF_LOG_ERROR,
+                                                "volume %s not present in "
+                                                "the cluster", volname);
+                                        goto out;
+                                }
+                                volid = gf_strdup (uuid_utoa (volinfo->volume_id));
+                                if (!volid) {
+                                        ret = -1;
+                                        goto out;
+                                }
+                                ret = dict_set_dynstr (dict, "vol-id", volid);
+                                if (ret) {
+                                        gf_log (THIS->name, GF_LOG_ERROR,
+                                                "Failed to set volume id in "
+                                                "dictionary");
+                                        goto out;
+                                }
                                 dict_copy (dict, req_dict);
                         }
                         break;