glusterd: Relocate rebalance sockfile

The defrag sockfile was moved from priv->workdir to
DEFAULT_VAR_RUN_DIRECTORY. The format for the new path of the defrag
sockfile is 'DEFAULT_VAR_RUN_DIRECTORY/gluster-rebalance-<vol-id>.sock'.

This was needed because the earlier location didn't have a fixed length
and could exceed UNIX_PATH_MAX characters. This could lead to the
rebalance process failing to start as the socket file could not be
created.

Also, for keeping backward compatiblity, glusterd_rebalance_rpc_create
will try both the new and old sockfile locations when attempting
reconnection.

Change-Id: I6740ea665de84ebce1ef7199c412f426de54e3d0
BUG: 1049726
Signed-off-by: Kaushal M <kaushal@redhat.com>
Reviewed-on: http://review.gluster.org/6616
Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>

3 files changed, 46 insertions, 8 deletions

diff --git a/xlators/mgmt/glusterd/src/glusterd-rebalance.c b/xlators/mgmt/glusterd/src/glusterd-rebalance.c
index daa8ddd1def..b2805613517 100644
--- a/xlators/mgmt/glusterd/src/glusterd-rebalance.c
+++ b/xlators/mgmt/glusterd/src/glusterd-rebalance.c
@@ -237,7 +237,7 @@ glusterd_handle_defrag_start (glusterd_volinfo_t *volinfo, char *op_errstr,
                 goto out;
         }
 
-        GLUSTERD_GET_DEFRAG_SOCK_FILE (sockfile, volinfo, priv);
+        GLUSTERD_GET_DEFRAG_SOCK_FILE (sockfile, volinfo);
         GLUSTERD_GET_DEFRAG_PID_FILE (pidfile, volinfo, priv);
         snprintf (logfile, PATH_MAX, "%s/%s-rebalance.log",
                     DEFAULT_LOG_FILE_DIRECTORY, volinfo->volname);
@@ -288,7 +288,7 @@ glusterd_handle_defrag_start (glusterd_volinfo_t *volinfo, char *op_errstr,
 
         sleep (5);
 
-        ret = glusterd_rebalance_rpc_create (volinfo);
+        ret = glusterd_rebalance_rpc_create (volinfo, _gf_false);
 
         //FIXME: this cbk is passed as NULL in all occurrences. May be
         //we never needed it.
@@ -302,13 +302,21 @@ out:
 
 
 int
-glusterd_rebalance_rpc_create (glusterd_volinfo_t *volinfo)
+glusterd_rebalance_rpc_create (glusterd_volinfo_t *volinfo,
+                               gf_boolean_t reconnect)
 {
         dict_t                  *options = NULL;
         char                     sockfile[PATH_MAX] = {0,};
         int                      ret = -1;
         glusterd_defrag_info_t  *defrag = volinfo->rebal.defrag;
-        glusterd_conf_t         *priv = THIS->private;
+        glusterd_conf_t         *priv = NULL;
+        xlator_t                *this = NULL;
+        struct stat             buf = {0,};
+
+        this = THIS;
+        GF_ASSERT (this);
+        priv = this->private;
+        GF_ASSERT (priv);
 
         //rebalance process is not started
         if (!defrag)
@@ -319,7 +327,30 @@ glusterd_rebalance_rpc_create (glusterd_volinfo_t *volinfo)
                 ret = 0;
                 goto out;
         }
-        GLUSTERD_GET_DEFRAG_SOCK_FILE (sockfile, volinfo, priv);
+        GLUSTERD_GET_DEFRAG_SOCK_FILE (sockfile, volinfo);
+        /* If reconnecting check if defrag sockfile exists in the new location
+         * in /var/run/ , if it does not try the old location
+         */
+        if (reconnect) {
+                ret = sys_stat (sockfile, &buf);
+                /* TODO: Remove this once we don't need backward compatability
+                 * with the older path
+                 */
+                if (ret && (errno == ENOENT)) {
+                        gf_log (this->name, GF_LOG_WARNING, "Rebalance sockfile "
+                                "%s does not exist. Trying old path.",
+                                sockfile);
+                        GLUSTERD_GET_DEFRAG_SOCK_FILE_OLD (sockfile, volinfo,
+                                                           priv);
+                        ret =sys_stat (sockfile, &buf);
+                        if (ret && (ENOENT == errno)) {
+                                gf_log (this->name, GF_LOG_ERROR, "Rebalance "
+                                        "sockfile %s does not exist.",
+                                        sockfile);
+                                goto out;
+                        }
+                }
+        }
 
         /* Setting frame-timeout to 10mins (600seconds).
          * Unix domain sockets ensures that the connection is reliable. The
diff --git a/xlators/mgmt/glusterd/src/glusterd-utils.c b/xlators/mgmt/glusterd/src/glusterd-utils.c
index a0c9691249a..4a88b5b3523 100644
--- a/xlators/mgmt/glusterd/src/glusterd-utils.c
+++ b/xlators/mgmt/glusterd/src/glusterd-utils.c
@@ -6949,7 +6949,7 @@ glusterd_volume_defrag_restart (glusterd_volinfo_t *volinfo, char *op_errstr,
         case GF_DEFRAG_STATUS_STARTED:
                 GLUSTERD_GET_DEFRAG_PID_FILE(pidfile, volinfo, priv);
                 if (gf_is_service_running (pidfile, &pid)) {
-                        glusterd_rebalance_rpc_create (volinfo);
+                        glusterd_rebalance_rpc_create (volinfo, _gf_true);
                         break;
                 }
         case GF_DEFRAG_STATUS_NOT_STARTED:
diff --git a/xlators/mgmt/glusterd/src/glusterd.h b/xlators/mgmt/glusterd/src/glusterd.h
index ab383ac1c9a..9b6e2fb3343 100644
--- a/xlators/mgmt/glusterd/src/glusterd.h
+++ b/xlators/mgmt/glusterd/src/glusterd.h
@@ -439,13 +439,19 @@ typedef ssize_t (*gd_serialize_t) (struct iovec outmsg, void *args);
                 snprintf (path, PATH_MAX, "%s/rebalance",vol_path);     \
         } while (0)
 
-#define GLUSTERD_GET_DEFRAG_SOCK_FILE(path, volinfo, priv) do {         \
+#define GLUSTERD_GET_DEFRAG_SOCK_FILE_OLD(path, volinfo, priv) do {     \
                 char defrag_path[PATH_MAX];                             \
                 GLUSTERD_GET_DEFRAG_DIR(defrag_path, volinfo, priv);    \
                 snprintf (path, PATH_MAX, "%s/%s.sock", defrag_path,    \
                            uuid_utoa(MY_UUID));                         \
         } while (0)
 
+#define GLUSTERD_GET_DEFRAG_SOCK_FILE(path, volinfo) do {                   \
+                snprintf (path, UNIX_PATH_MAX, DEFAULT_VAR_RUN_DIRECTORY    \
+                          "/gluster-rebalance-%s.sock",                     \
+                           uuid_utoa(volinfo->volume_id));                  \
+        } while (0)
+
 #define GLUSTERD_GET_DEFRAG_PID_FILE(path, volinfo, priv) do {          \
                 char defrag_path[PATH_MAX];                             \
                 GLUSTERD_GET_DEFRAG_DIR(defrag_path, volinfo, priv);    \
@@ -727,7 +733,8 @@ int glusterd_handle_defrag_start (glusterd_volinfo_t *volinfo, char *op_errstr,
                                   size_t len, int cmd, defrag_cbk_fn_t cbk,
                                   glusterd_op_t op);
 int
-glusterd_rebalance_rpc_create (glusterd_volinfo_t *volinfo);
+glusterd_rebalance_rpc_create (glusterd_volinfo_t *volinfo,
+                               gf_boolean_t reconnect);
 
 int glusterd_handle_cli_heal_volume (rpcsvc_request_t *req);