cluster/afr: Send truncate on arbiter brick from SHD

Problem:
In an arbiter volume configuration SHD will not send any writes onto the arbiter
brick even if there is data pending marker for the arbiter brick. If we have a
arbiter setup on the geo-rep master and there are data pending markers for the files
on arbiter brick, SHD will not mark any data changelog during healing. While syncing
the data from master to slave, if the arbiter-brick is considered as ACTIVE, then
there is a chance that slave will miss out some data. If the arbiter brick is being
newly added or replaced there is a chance of slave missing all the data during sync.

Fix:
If there is data pending marker for the arbiter brick, send truncate on the arbiter
brick during heal, so that it will record truncate as the data transaction in changelog.

Change-Id: I3242ba6cea6da495c418ef860d9c3359c5459dec
fixes: bz#1687687
Signed-off-by: karthik-us <ksubrahm@redhat.com>

1 files changed, 13 insertions, 15 deletions

diff --git a/xlators/cluster/afr/src/afr-self-heal-data.c b/xlators/cluster/afr/src/afr-self-heal-data.c
index 9b296b9ad23..2aed60baa1a 100644
--- a/xlators/cluster/afr/src/afr-self-heal-data.c
+++ b/xlators/cluster/afr/src/afr-self-heal-data.c
@@ -382,17 +382,18 @@ __afr_selfheal_truncate_sinks(call_frame_t *frame, xlator_t *this, fd_t *fd,
 {
     afr_local_t *local = NULL;
     afr_private_t *priv = NULL;
-    unsigned char arbiter_sink_status = 0;
     int i = 0;
 
     local = frame->local;
     priv = this->private;
 
-    if (priv->arbiter_count) {
-        arbiter_sink_status = healed_sinks[ARBITER_BRICK_INDEX];
-        healed_sinks[ARBITER_BRICK_INDEX] = 0;
-    }
-
+    /* This will send truncate on the arbiter brick as well if it is marked as
+     * sink. If changelog is enabled on the volume it captures truncate as a
+     * data transactions on the arbiter brick. This will help geo-rep to
+     * properly sync the data from master to slave if arbiter is the ACTIVE
+     * brick during syncing and which had got some entries healed for data as
+     * part of self heal.
+     */
     AFR_ONLIST(healed_sinks, frame, afr_sh_generic_fop_cbk, ftruncate, fd, size,
                NULL);
 
@@ -403,8 +404,6 @@ __afr_selfheal_truncate_sinks(call_frame_t *frame, xlator_t *this, fd_t *fd,
             */
             healed_sinks[i] = 0;
 
-    if (arbiter_sink_status)
-        healed_sinks[ARBITER_BRICK_INDEX] = arbiter_sink_status;
     return 0;
 }
 
@@ -700,19 +699,18 @@ __afr_selfheal_data(call_frame_t *frame, xlator_t *this, fd_t *fd,
             goto unlock;
         }
 
-        if (priv->arbiter_count &&
-            AFR_COUNT(healed_sinks, priv->child_count) == 1 &&
-            healed_sinks[ARBITER_BRICK_INDEX]) {
-            is_arbiter_the_only_sink = _gf_true;
-            goto restore_time;
-        }
-
         ret = __afr_selfheal_truncate_sinks(
             frame, this, fd, healed_sinks,
             locked_replies[source].poststat.ia_size);
         if (ret < 0)
             goto unlock;
 
+        if (priv->arbiter_count &&
+            AFR_COUNT(healed_sinks, priv->child_count) == 1 &&
+            healed_sinks[ARBITER_BRICK_INDEX]) {
+            is_arbiter_the_only_sink = _gf_true;
+            goto restore_time;
+        }
         ret = 0;
     }
 unlock: