cluster/afr: Use parallel dir scan functionality

BUG: 1221737
Change-Id: I0ed71a72f0e33bd733723e00a01cf28378c5534e
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>
Reviewed-on: http://review.gluster.org/13755
NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org>
CentOS-regression: Gluster Build System <jenkins@build.gluster.com>
Smoke: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Jeff Darcy <jdarcy@redhat.com>

4 files changed, 59 insertions, 13 deletions

diff --git a/xlators/cluster/afr/src/afr-self-heald.c b/xlators/cluster/afr/src/afr-self-heald.c
index 9439bba63ff..1ae4f18e764 100644
--- a/xlators/cluster/afr/src/afr-self-heald.c
+++ b/xlators/cluster/afr/src/afr-self-heald.c
@@ -315,14 +315,18 @@ afr_shd_selfheal (struct subvol_healer *healer, int child, uuid_t gfid)
 
 	ret = afr_selfheal (this, gfid);
 
-	if (ret == -EIO) {
-		eh = shd->split_brain;
-		crawl_event->split_brain_count++;
-	} else if (ret < 0) {
-		crawl_event->heal_failed_count++;
-	} else if (ret == 0) {
-		crawl_event->healed_count++;
-	}
+        LOCK (&priv->lock);
+        {
+                if (ret == -EIO) {
+                        eh = shd->split_brain;
+                        crawl_event->split_brain_count++;
+                } else if (ret < 0) {
+                        crawl_event->heal_failed_count++;
+                } else if (ret == 0) {
+                        crawl_event->healed_count++;
+                }
+        }
+        UNLOCK (&priv->lock);
 
 	if (eh) {
 		shd_event = GF_CALLOC (1, sizeof(*shd_event),
@@ -425,6 +429,7 @@ afr_shd_index_sweep (struct subvol_healer *healer, char *vgfid)
 	afr_private_t *priv   = NULL;
 	int           ret     = 0;
 	xlator_t      *subvol = NULL;
+	dict_t        *xdata  = NULL;
 
 	priv = healer->this->private;
 	subvol = priv->children[healer->subvol];
@@ -434,17 +439,28 @@ afr_shd_index_sweep (struct subvol_healer *healer, char *vgfid)
 	        gf_msg (healer->this->name, GF_LOG_WARNING,
                         0, AFR_MSG_INDEX_DIR_GET_FAILED,
 		        "unable to get index-dir on %s", subvol->name);
-		return -errno;
+		ret = -errno;
+	        goto out;
 	}
 
-        ret = syncop_dir_scan (subvol, &loc, GF_CLIENT_PID_SELF_HEALD,
-                               healer, afr_shd_index_heal);
+        xdata = dict_new ();
+        if (!xdata || dict_set_int32 (xdata, "get-gfid-type", 1)) {
+                ret = -ENOMEM;
+                goto out;
+        }
 
-        loc_wipe (&loc);
+        ret = syncop_mt_dir_scan (subvol, &loc, GF_CLIENT_PID_SELF_HEALD,
+                                  healer, afr_shd_index_heal, xdata,
+                                 priv->shd.max_threads, priv->shd.wait_qlength);
 
         if (ret == 0)
                 ret = healer->crawl_event.healed_count;
 
+out:
+        loc_wipe (&loc);
+
+        if (xdata)
+                dict_unref (xdata);
 	return ret;
 }
 
diff --git a/xlators/cluster/afr/src/afr-self-heald.h b/xlators/cluster/afr/src/afr-self-heald.h
index 58b088e4cd7..f591515669c 100644
--- a/xlators/cluster/afr/src/afr-self-heald.h
+++ b/xlators/cluster/afr/src/afr-self-heald.h
@@ -56,6 +56,8 @@ typedef struct {
 
         eh_t                    *split_brain;
         eh_t                    **statistics;
+        uint32_t                max_threads;
+        uint32_t                wait_qlength;
 } afr_self_heald_t;
 
 
diff --git a/xlators/cluster/afr/src/afr.c b/xlators/cluster/afr/src/afr.c
index 2c7e69aae3b..6e59fd46328 100644
--- a/xlators/cluster/afr/src/afr.c
+++ b/xlators/cluster/afr/src/afr.c
@@ -217,6 +217,12 @@ reconfigure (xlator_t *this, dict_t *options)
         GF_OPTION_RECONF ("consistent-metadata", priv->consistent_metadata,
                           options, bool, out);
 
+        GF_OPTION_RECONF ("shd-max-threads", priv->shd.max_threads,
+                          options, uint32, out);
+
+        GF_OPTION_RECONF ("shd-wait-qlength", priv->shd.wait_qlength,
+                          options, uint32, out);
+
         priv->did_discovery = _gf_false;
 
         ret = 0;
@@ -329,6 +335,11 @@ init (xlator_t *this)
                         fav_child->name, fav_child->name);
         }
 
+        GF_OPTION_INIT ("shd-max-threads", priv->shd.max_threads,
+                         uint32, out);
+
+        GF_OPTION_INIT ("shd-wait-qlength", priv->shd.wait_qlength,
+                         uint32, out);
 
         GF_OPTION_INIT ("background-self-heal-count",
                         priv->background_self_heal_count, uint32, out);
@@ -850,5 +861,23 @@ struct volume_options options[] = {
           .type = GF_OPTION_TYPE_INT,
           .description = "subset of child_count. Has to be 0 or 1."
         },
+        { .key   = {"shd-max-threads"},
+          .type  = GF_OPTION_TYPE_INT,
+          .min   = 1,
+          .max   = 64,
+          .default_value = "1",
+           .description = "Maximum number of threads SHD can use per local "
+                          "brick.  This can substantially lower heal times, "
+                          "but can also crush your bricks if you don't have "
+                          "the storage hardware to support this."
+        },
+        { .key   = {"shd-wait-qlength"},
+          .type  = GF_OPTION_TYPE_INT,
+          .min   = 1,
+          .max   = 655536,
+          .default_value = "1024",
+           .description = "This option can be used to control number of heals"
+                          " that can wait in SHD per subvolume",
+        },
         { .key  = {NULL} },
 };
diff --git a/xlators/cluster/afr/src/afr.h b/xlators/cluster/afr/src/afr.h
index 56682b47a56..7a99e70c80e 100644
--- a/xlators/cluster/afr/src/afr.h
+++ b/xlators/cluster/afr/src/afr.h
@@ -139,7 +139,6 @@ typedef struct _afr_private {
 	/* pump dependencies */
 	void                   *pump_private;
 	gf_boolean_t           use_afr_in_pump;
-
 } afr_private_t;