cluster/tier: account for reordered layouts

This is a backport of 11092

> For a tiered volume the cold subvolume is always at a fixed
> position in the graph. DHT's layout array, on the other hand,
> may have the cold subvolume in either the first or second
> index, therefore code cannot make any assumptions. The fix
> searches the layout for the correct position dynamically
> rather than statically.

> The bug manifested itself in NFS, in which a newly attached
> subvolume had not received an existing directory. This case
> is a "stale entry" and marked as such in the layout for
> that directory.  The code did not see this, because it
> looked at the wrong index in the layout array.

> The fix also adds the check for decomissioned bricks, and
> fixes a problem in detach tier related to starting the
> rebalance process: we never received the right defrag
> command and it did not get directed to the tier translator.

> Change-Id: I77cdf9fbb0a777640c98003188565a79be9d0b56
> BUG: 1214289
> Signed-off-by: Dan Lambright <dlambrig@redhat.com>

Change-Id: Idb2eec9ba25812f41de7f960a0314c92341d6b5d
BUG: 1259081
Signed-off-by: Dan Lambright <dlambrig@redhat.com>
Reviewed-on: http://review.gluster.org/12086
Tested-by: NetBSD Build System <jenkins@build.gluster.org>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: mohammed rafi  kc <rkavunga@redhat.com>

1 files changed, 30 insertions, 13 deletions

diff --git a/xlators/cluster/dht/src/tier.c b/xlators/cluster/dht/src/tier.c
index 59e260af2dc..e996c22ea1d 100644
--- a/xlators/cluster/dht/src/tier.c
+++ b/xlators/cluster/dht/src/tier.c
@@ -947,7 +947,8 @@ tier_migration_needed (xlator_t *this)
 
         defrag = conf->defrag;
 
-        if (defrag->cmd == GF_DEFRAG_CMD_START_TIER)
+        if ((defrag->cmd == GF_DEFRAG_CMD_START_TIER) ||
+            (defrag->cmd == GF_DEFRAG_CMD_START_DETACH_TIER))
                 ret = 1;
 out:
         return ret;
@@ -989,9 +990,11 @@ tier_search (xlator_t *this, dht_layout_t *layout, const char *name)
 {
         xlator_t                *subvol = NULL;
         void                    *value;
-        int                      search_first_subvol = 0;
+        int                      search_subvol = 0;
         dht_conf_t              *conf   = NULL;
         gf_defrag_info_t        *defrag = NULL;
+        int                      layout_cold = 0;
+        int                      layout_hot = 1;
 
         GF_VALIDATE_OR_GOTO("tier", this, out);
         GF_VALIDATE_OR_GOTO(this->name, layout, out);
@@ -1000,28 +1003,42 @@ tier_search (xlator_t *this, dht_layout_t *layout, const char *name)
 
         conf = this->private;
 
+        /* The first subvolume in the graph is always cold. */
+        /* Find the position of the cold subvolume in the layout. */
+        layout_cold = 0;
+        layout_hot = 1;
+        if (conf->subvolumes[0] != layout->list[0].xlator) {
+                layout_cold = 1;
+                layout_hot = 0;
+        }
+
+        search_subvol = layout_hot;
+
         defrag = conf->defrag;
         if (defrag && defrag->cmd == GF_DEFRAG_CMD_START_DETACH_TIER)
-                search_first_subvol = 1;
+                search_subvol = layout_cold;
 
+        /* "decommission_subvols_cnt" can only be non-zero on detach. */
+        /* This will change once brick add/remove is supported for */
+        /* tiered volumes. */
+        else if (conf->decommission_subvols_cnt) {
+                search_subvol = layout_cold;
+        }
         else if (!dict_get_ptr (this->options, "rule", &value) &&
-                 !strcmp(layout->list[0].xlator->name, value)) {
-                search_first_subvol = 1;
+                 !strcmp(layout->list[layout_cold].xlator->name, value)) {
+                search_subvol = layout_cold;
         }
 
-        if ((layout->list[0].err > 0) && (layout->list[0].err != ENOTCONN))
-                search_first_subvol = 0;
+        if ((layout->list[search_subvol].err > 0) &&
+            (layout->list[search_subvol].err != ENOTCONN))
+                search_subvol = layout_cold;
 
-        if (search_first_subvol)
-                subvol = layout->list[0].xlator;
-        else
-                subvol = layout->list[1].xlator;
+        subvol = layout->list[search_subvol].xlator;
+ out:
 
-out:
         return subvol;
 }
 
-
 dht_methods_t tier_methods = {
         .migration_get_dst_subvol = tier_migration_get_dst,
         .migration_other = tier_start,