geo-rep: rename handling in dht volume(changelog changes)

Background:

Glusterfs changelogs are stored in each brick, which records the changes
happened in that brick. Georep will run in all the nodes of master and processes
changelogs "independently".  Processing changelogs is in brick level, but
all the fops will be replayed on "slave mount" point.

Problem:

With a DHT volume, in changelog "internal fops" are NOT recorded.
For Rename case, Rename is recorded in "hashed" brick changelog.
(DHT's internal fops like creating linkto file, unlink is NOT recorded).
This lead us to inconsistent rename operations.

For example,
Distribute volume created with Two bricks B1, B2.

//Consider master volume mounted @ /mnt/master and following operations
executed:
cd /mnt/master
touch f1 // f1 falls on B1 Hash
mv f1 f2 // f2 falls on B2 Hash

// Here, Changelogs are recorded as below:
@B1
CREATE f1

@B2
RENAME f1 f2

Here, race exists between Brick B1 and B2, say B2 will get executed first.
Source file f1 itself is "NOT PRESENT", so it will go ahead and create
f2 (Current implementation).

We have this problem When rename falls in another brick and
file is unlinked in Master.

Similar kind of issue exists in following case too(multiple rename):
CREATE f1
RENAME f1 f2
RENAME f2 f1

Solution:

Instead of carrying out "changelogging" at "HASHED volume",
carry out  at the "CACHED volume".
This way we have rename operations carried out where actual files are present.

So,Changelog recorded as :
@B1
CREATE f1
RENAME f1 f2

Note:
This patch is dependent on dht changes from this patch.
http://review.gluster.org/10410/
changelog related changes are separated out for review.

In changelog, xdata passed from DHT is considered as :

1. In case of unlink (internal operation as part of rename), xdata value
is set , it is considered as RENAME and recorded accordingly.
2. In case of rename (Hash and Cache different), xdata value is NOT
set, recording rename operation is SKIPPED.

BUG: 1141379
Change-Id: Ifca675e6d4ef8c4e3b7ef4a7ec85de8b3a38dc08
Signed-off-by: Saravanakumar Arumugam <sarumuga@redhat.com>
Reviewed-on: http://review.gluster.org/10220
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Tested-by: NetBSD Build System
Reviewed-by: Kotresh HR <khiremat@redhat.com>

1 files changed, 77 insertions, 38 deletions

diff --git a/xlators/features/changelog/src/changelog.c b/xlators/features/changelog/src/changelog.c
index 53c0cf85728..b07009d3f8f 100644
--- a/xlators/features/changelog/src/changelog.c
+++ b/xlators/features/changelog/src/changelog.c
@@ -212,40 +212,76 @@ int32_t
 changelog_unlink (call_frame_t *frame, xlator_t *this,
                   loc_t *loc, int xflags, dict_t *xdata)
 {
-        size_t                  xtra_len                = 0;
-        changelog_priv_t       *priv                    = NULL;
-        changelog_opt_t        *co                      = NULL;
-        call_stub_t            *stub                    = NULL;
-        struct list_head        queue                   = {0, };
-        gf_boolean_t            barrier_enabled         = _gf_false;
+        size_t                       xtra_len        = 0;
+        changelog_priv_t             *priv           = NULL;
+        changelog_opt_t              *co             = NULL;
+        call_stub_t                  *stub           = NULL;
+        struct list_head             queue           = {0, };
+        gf_boolean_t                 barrier_enabled = _gf_false;
+        dht_changelog_rename_info_t  *info           = NULL;
+        int                          ret             = 0;
+        char                         old_name[NAME_MAX] = {0};
+        char                         new_name[NAME_MAX] = {0};
+        char                         *nname             = NULL;
 
         INIT_LIST_HEAD (&queue);
-
         priv = this->private;
-        CHANGELOG_NOT_ACTIVE_THEN_GOTO (frame, priv, wind);
-        CHANGELOG_IF_INTERNAL_FOP_THEN_GOTO (frame, xdata, wind);
 
-        CHANGELOG_INIT_NOCHECK (this, frame->local, NULL, loc->inode->gfid, 2);
-
-        co = changelog_get_usable_buffer (frame->local);
-        if (!co)
-                goto wind;
+        CHANGELOG_NOT_ACTIVE_THEN_GOTO (frame, priv, wind);
 
-        CHANGLOG_FILL_FOP_NUMBER (co, frame->root->op, fop_fn, xtra_len);
+        ret = dict_get_bin (xdata, DHT_CHANGELOG_RENAME_OP_KEY, (void **)&info);
+        if (!ret) {     /* special case: unlink considered as rename */
+                /* 3 == fop + oldloc + newloc */
+                CHANGELOG_INIT_NOCHECK (this, frame->local,
+                                        NULL, loc->inode->gfid, 3);
+
+                co = changelog_get_usable_buffer (frame->local);
+                if (!co)
+                        goto wind;
+
+                CHANGLOG_FILL_FOP_NUMBER (co, GF_FOP_RENAME, fop_fn, xtra_len);
+
+                co++;
+                strncpy (old_name, info->buffer, info->oldname_len);
+                CHANGELOG_FILL_ENTRY (co, info->old_pargfid, old_name,
+                                      entry_fn, entry_free_fn, xtra_len, wind);
+
+                co++;
+                /* new name resides just after old name */
+                nname = info->buffer + info->oldname_len;
+                strncpy (new_name, nname, info->newname_len);
+                CHANGELOG_FILL_ENTRY (co, info->new_pargfid, new_name,
+                                      entry_fn, entry_free_fn, xtra_len, wind);
+
+                changelog_set_usable_record_and_length (frame->local,
+                                                        xtra_len, 3);
+        } else {        /* default unlink */
+                CHANGELOG_IF_INTERNAL_FOP_THEN_GOTO (frame, xdata, wind);
+                CHANGELOG_INIT_NOCHECK (this, frame->local, NULL,
+                                                        loc->inode->gfid, 2);
+
+                co = changelog_get_usable_buffer (frame->local);
+                if (!co)
+                        goto wind;
+
+                CHANGLOG_FILL_FOP_NUMBER (co, frame->root->op,
+                                                fop_fn, xtra_len);
+
+                co++;
+                if (priv->capture_del_path) {
+                        CHANGELOG_FILL_ENTRY_DIR_PATH (co, loc->pargfid,
+                                     loc->name, del_entry_fn, del_entry_free_fn,
+                                     xtra_len, wind, _gf_true);
+                } else {
+                        CHANGELOG_FILL_ENTRY_DIR_PATH (co, loc->pargfid,
+                                     loc->name, del_entry_fn, del_entry_free_fn,
+                                     xtra_len, wind, _gf_false);
+                }
 
-        co++;
-        if (priv->capture_del_path) {
-                CHANGELOG_FILL_ENTRY_DIR_PATH (co, loc->pargfid, loc->name,
-                                               del_entry_fn, del_entry_free_fn,
-                                              xtra_len, wind, _gf_true);
-        } else {
-                CHANGELOG_FILL_ENTRY_DIR_PATH (co, loc->pargfid, loc->name,
-                                               del_entry_fn, del_entry_free_fn,
-                                              xtra_len, wind, _gf_false);
+                changelog_set_usable_record_and_length (frame->local,
+                                                        xtra_len, 2);
         }
 
-        changelog_set_usable_record_and_length (frame->local, xtra_len, 2);
-
 /* changelog barrier */
         LOCK (&priv->lock);
         {
@@ -295,16 +331,13 @@ changelog_rename_cbk (call_frame_t *frame,
                       struct iatt *postoldparent, struct iatt *prenewparent,
                       struct iatt *postnewparent, dict_t *xdata)
 {
-        changelog_priv_t  *priv  = NULL;
-        changelog_local_t *local = NULL;
+        changelog_priv_t                  *priv  = NULL;
+        changelog_local_t                 *local = NULL;
 
         priv  = this->private;
         local = frame->local;
-
         CHANGELOG_COND_GOTO (priv, ((op_ret < 0) || !local), unwind);
-
         changelog_update (this, priv, local, CHANGELOG_TYPE_ENTRY);
-
  unwind:
         changelog_dec_fop_cnt (this, priv, local);
         CHANGELOG_STACK_UNWIND (rename, frame, op_ret, op_errno,
@@ -313,7 +346,6 @@ changelog_rename_cbk (call_frame_t *frame,
         return 0;
 }
 
-
 int32_t
 changelog_rename_resume (call_frame_t *frame, xlator_t *this,
                          loc_t *oldloc, loc_t *newloc, dict_t *xdata)
@@ -334,18 +366,25 @@ int32_t
 changelog_rename (call_frame_t *frame, xlator_t *this,
                   loc_t *oldloc, loc_t *newloc, dict_t *xdata)
 {
-        size_t                  xtra_len                = 0;
-        changelog_priv_t       *priv                    = NULL;
-        changelog_opt_t        *co                      = NULL;
-        call_stub_t            *stub                    = NULL;
-        struct list_head        queue                   = {0, };
-        gf_boolean_t            barrier_enabled         = _gf_false;
+        size_t               xtra_len        = 0;
+        changelog_priv_t     *priv           = NULL;
+        changelog_opt_t      *co             = NULL;
+        call_stub_t          *stub           = NULL;
+        struct list_head     queue           = {0, };
+        gf_boolean_t         barrier_enabled = _gf_false;
+        dht_changelog_rename_info_t  *info   = NULL;
+        int                  ret             = 0;
 
         INIT_LIST_HEAD (&queue);
 
         priv = this->private;
         CHANGELOG_NOT_ACTIVE_THEN_GOTO (frame, priv, wind);
 
+        ret = dict_get_bin (xdata, DHT_CHANGELOG_RENAME_OP_KEY, (void **)&info);
+        if (ret)  {      /* xdata "NOT" set, Special rename => avoid logging */
+               goto wind;
+        }
+
         /* 3 == fop + oldloc + newloc */
         CHANGELOG_INIT_NOCHECK (this, frame->local,
                                 NULL, oldloc->inode->gfid, 3);