cluster/dht: Heal dir uid/gid

Identify mismatching uid/gid in lookup, and trigger a syncop
heal. uid/gid of subvol with latest ctime is trusted (local->prebuf).

Change-Id: Ib5c4bc438e7f4b1f33080e73593f40f400e997f0
BUG: 862967
Signed-off-by: shishir gowda <sgowda@redhat.com>
Reviewed-on: http://review.gluster.org/3964
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>

5 files changed, 162 insertions, 1 deletions

diff --git a/tests/bugs/bug-862967.t b/tests/bugs/bug-862967.t
new file mode 100644
index 00000000000..73f5826fa47
--- /dev/null
+++ b/tests/bugs/bug-862967.t
@@ -0,0 +1,57 @@
+#!/bin/bash
+
+. $(dirname $0)/../include.rc
+
+cleanup;
+
+function uid_gid_compare()
+{
+        val=1
+
+        if [ "$1" == "$3" ]
+        then
+                if [ "$2" == "$4" ]
+                        then
+                                val=0
+                fi
+        fi
+        echo "$val"
+}
+
+BRICK_COUNT=3
+
+TEST glusterd
+TEST pidof glusterd
+
+TEST $CLI volume create $V0 $H0:$B0/${V0}0 $H0:$B0/${V0}1 $H0:$B0/${V0}2
+TEST $CLI volume set $V0 stat-prefetch off
+TEST $CLI volume start $V0
+
+## Mount FUSE
+TEST glusterfs --attribute-timeout=0 --entry-timeout=0 --gid-timeout=-1 -s $H0 --volfile-id $V0 $M0;
+
+# change dir permissions
+mkdir $M0/dir;
+chown 1:1 $M0/dir;
+
+# Kill a brick process
+
+kill -9 `cat /var/lib/glusterd/vols/$V0/run/$H0-d-backends-${V0}1.pid`;
+# change dir ownership
+NEW_UID=36;
+NEW_GID=36;
+chown $NEW_UID:$NEW_GID $M0/dir;
+
+# bring the brick back up
+TEST $CLI volume start $V0 force
+
+sleep 10;
+
+ls -l $M0/dir;
+
+# check if uid/gid is healed on backend brick which was taken down
+BACKEND_UID=`stat --printf=%u $B0/${V0}1/dir`;
+BACKEND_GID=`stat --printf=%g $B0/${V0}1/dir`;
+
+
+EXPECT "0" uid_gid_compare $NEW_UID $NEW_GID $BACKEND_UID $BACKEND_GID
diff --git a/xlators/cluster/dht/src/dht-common.c b/xlators/cluster/dht/src/dht-common.c
index 98560685bce..0925b6d2471 100644
--- a/xlators/cluster/dht/src/dht-common.c
+++ b/xlators/cluster/dht/src/dht-common.c
@@ -525,6 +525,8 @@ dht_revalidate_cbk (call_frame_t *frame, void *cookie, xlator_t *this,
         int           ret  = -1;
         int           is_dir = 0;
         int           is_linkfile = 0;
+        call_frame_t *copy          = NULL;
+        dht_local_t  *copy_local    = NULL;
 
         GF_VALIDATE_OR_GOTO ("dht", frame, err);
         GF_VALIDATE_OR_GOTO ("dht", this, err);
@@ -595,6 +597,23 @@ dht_revalidate_cbk (call_frame_t *frame, void *cookie, xlator_t *this,
                 }
 
                 if (is_dir) {
+                        ret = dht_dir_has_layout (xattr);
+                        if (ret >= 0) {
+                                if (is_greater_time(local->stbuf.ia_ctime,
+                                                    local->stbuf.ia_ctime_nsec,
+                                                    stbuf->ia_ctime,
+                                                    stbuf->ia_ctime_nsec)) {
+                                        local->prebuf.ia_gid = stbuf->ia_gid;
+                                        local->prebuf.ia_uid = stbuf->ia_uid;
+                                }
+                        }
+                        if (local->stbuf.ia_type != IA_INVAL)
+                        {
+                                if ((local->stbuf.ia_gid != stbuf->ia_gid) ||
+                                    (local->stbuf.ia_uid != stbuf->ia_uid)) {
+                                        local->need_selfheal = 1;
+                                }
+                        }
                         ret = dht_layout_dir_mismatch (this, layout,
                                                        prev->this, &local->loc,
                                                        xattr);
@@ -633,7 +652,28 @@ out:
                     && (conf && conf->unhashed_sticky_bit)) {
                         local->stbuf.ia_prot.sticky = 1;
                 }
-                if (local->layout_mismatch) {
+                if (local->need_selfheal) {
+                        local->need_selfheal = 0;
+                        uuid_copy (local->gfid, local->stbuf.ia_gfid);
+                        local->stbuf.ia_gid = local->prebuf.ia_gid;
+                        local->stbuf.ia_uid = local->prebuf.ia_uid;
+                        copy = create_frame (this, this->ctx->pool);
+                        if (copy) {
+                                copy_local = dht_local_init (copy, &local->loc,
+                                                             NULL, 0);
+                                if (!copy_local)
+                                        goto cont;
+                                copy_local->stbuf = local->stbuf;
+                                copy->local = copy_local;
+                                FRAME_SU_DO (copy, dht_local_t);
+                                ret = synctask_new (this->ctx->env,
+                                                    dht_dir_attr_heal,
+                                                    dht_dir_attr_heal_done,
+                                                    copy, copy);
+                        }
+                }
+cont:
+		if (local->layout_mismatch) {
                         /* Found layout mismatch in the directory, need to
                            fix this in the inode context */
                         dht_layout_unref (this, local->layout);
diff --git a/xlators/cluster/dht/src/dht-common.h b/xlators/cluster/dht/src/dht-common.h
index 0f2df514fb5..1a69feba607 100644
--- a/xlators/cluster/dht/src/dht-common.h
+++ b/xlators/cluster/dht/src/dht-common.h
@@ -370,6 +370,7 @@ typedef enum {
                 UNLOCK (&inode->lock);                                  \
         } while (0)
 
+#define is_greater_time(a, an, b, bn) (((a) < (b)) || (((a) == (b)) && ((an) < (bn))))
 dht_layout_t                            *dht_layout_new (xlator_t *this, int cnt);
 dht_layout_t                            *dht_layout_get (xlator_t *this, inode_t *inode);
 dht_layout_t                            *dht_layout_for_subvol (xlator_t *this, xlator_t *subvol);
@@ -715,4 +716,10 @@ dht_inode_ctx_time_update (inode_t *inode, xlator_t *this, struct iatt *stat,
 
 int dht_inode_ctx_get (inode_t *inode, xlator_t *this, dht_inode_ctx_t **ctx);
 int dht_inode_ctx_set (inode_t *inode, xlator_t *this, dht_inode_ctx_t *ctx);
+int
+dht_dir_attr_heal (void *data);
+int
+dht_dir_attr_heal_done (int ret, call_frame_t *sync_frame, void *data);
+int
+dht_dir_has_layout (dict_t *xattr);
 #endif/* _DHT_H */
diff --git a/xlators/cluster/dht/src/dht-layout.c b/xlators/cluster/dht/src/dht-layout.c
index 19807607758..993f6166a12 100644
--- a/xlators/cluster/dht/src/dht-layout.c
+++ b/xlators/cluster/dht/src/dht-layout.c
@@ -615,6 +615,16 @@ out:
         return ret;
 }
 
+int
+dht_dir_has_layout (dict_t *xattr)
+{
+
+        void     *disk_layout_raw = NULL;
+
+        return dict_get_ptr (xattr, "trusted.glusterfs.dht",
+                             &disk_layout_raw);
+
+}
 
 int
 dht_layout_dir_mismatch (xlator_t *this, dht_layout_t *layout, xlator_t *subvol,
diff --git a/xlators/cluster/dht/src/dht-selfheal.c b/xlators/cluster/dht/src/dht-selfheal.c
index 8ac5a9e9e44..7a5d4d116ad 100644
--- a/xlators/cluster/dht/src/dht-selfheal.c
+++ b/xlators/cluster/dht/src/dht-selfheal.c
@@ -858,3 +858,50 @@ dht_selfheal_restore (call_frame_t *frame, dht_selfheal_dir_cbk_t dir_cbk,
 
         return ret;
 }
+
+int
+dht_dir_attr_heal (void *data)
+{
+        call_frame_t    *frame = NULL;
+        dht_local_t     *local = NULL;
+        xlator_t        *subvol = NULL;
+        xlator_t        *this  = NULL;
+        dht_conf_t      *conf  = NULL;
+        int              call_cnt = 0;
+        int              ret   = -1;
+        int              i     = 0;
+
+        GF_VALIDATE_OR_GOTO ("dht", data, out);
+
+        frame = data;
+        local = frame->local;
+        this = frame->this;
+        GF_VALIDATE_OR_GOTO ("dht", this, out);
+        GF_VALIDATE_OR_GOTO ("dht", local, out);
+        conf = this->private;
+        GF_VALIDATE_OR_GOTO ("dht", conf, out);
+
+        call_cnt = conf->subvolume_cnt;
+
+        for (i = 0; i < call_cnt; i++) {
+                subvol = conf->subvolumes[i];
+                if (!subvol || (subvol == dht_first_up_subvol (this)))
+                        continue;
+                ret = syncop_setattr (subvol, &local->loc, &local->stbuf,
+                                      (GF_SET_ATTR_UID | GF_SET_ATTR_GID),
+                                      NULL, NULL);
+                if (ret)
+                        gf_log ("dht", GF_LOG_ERROR, "Failed to set uid/gid on"
+                                " %s on %s subvol (%s)", local->loc.path,
+                                subvol->name, strerror (errno));
+        }
+out:
+        return 0;
+}
+
+int
+dht_dir_attr_heal_done (int ret, call_frame_t *sync_frame, void *data)
+{
+        DHT_STACK_DESTROY (sync_frame);
+        return 0;
+}