cluster/ec: Forced unlock when lock contention is detected

EC uses an eager lock mechanism to optimize multiple read/write
requests on the same entry or inode. This increases performance
but can have adverse results when other clients try to access the
same entry/inode.

To solve this, this patch adds a functionality to detect when this
happens and force an earlier release to not block other clients.

The method consists on requesting GF_GLUSTERFS_INODELK_COUNT and
GF_GLUSTERFS_ENTRYLK_COUNT for all fops that take a lock. When this
count is greater than one, the lock is marked to be released. All
fops already waiting for this lock will be executed normally before
releasing the lock, but new requests that also require it will be
blocked and restarted after the lock has been released and reacquired
again.

Another problem was that some operations did correctly lock the
parent of an entry when needed, but got the size and version xattrs
from the entry instead of the parent.

This patch solves this problem by binding all queries of size and
version to each lock and replacing all entrylk calls by inodelk ones
to remove concurrent updates on directory metadata.  This also allows
rename to correctly update source and destination directories.

Change-Id: I2df0b22bc6f407d49f3cbf0733b0720015bacfbd
BUG: 1165041
Signed-off-by: Xavier Hernandez <xhernandez@datalab.es>
Reviewed-on: http://review.gluster.org/10852
Tested-by: NetBSD Build System
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>

1 files changed, 16 insertions, 19 deletions

diff --git a/xlators/cluster/ec/src/ec-combine.c b/xlators/cluster/ec/src/ec-combine.c
index 9d4a18999f1..4617a0430f1 100644
--- a/xlators/cluster/ec/src/ec-combine.c
+++ b/xlators/cluster/ec/src/ec-combine.c
@@ -171,8 +171,10 @@ void ec_iatt_rebuild(ec_t * ec, struct iatt * iatt, int32_t count,
 gf_boolean_t
 ec_xattr_match (dict_t *dict, char *key, data_t *value, void *arg)
 {
-        if (fnmatch(GF_XATTR_STIME_PATTERN, key, 0) == 0)
+        if ((fnmatch(GF_XATTR_STIME_PATTERN, key, 0) == 0) ||
+            (strcmp(key, GLUSTERFS_OPEN_FD_COUNT) == 0)) {
                 return _gf_false;
+        }
 
         return _gf_true;
 }
@@ -185,6 +187,8 @@ ec_value_ignore (char *key)
             (strcmp(key, GF_XATTR_USER_PATHINFO_KEY) == 0) ||
             (strcmp(key, GF_XATTR_LOCKINFO_KEY) == 0) ||
             (strcmp(key, GLUSTERFS_OPEN_FD_COUNT) == 0) ||
+            (strcmp(key, GLUSTERFS_INODELK_COUNT) == 0) ||
+            (strcmp(key, GLUSTERFS_ENTRYLK_COUNT) == 0) ||
             (strncmp(key, GF_XATTR_CLRLK_CMD,
                      strlen (GF_XATTR_CLRLK_CMD)) == 0) ||
             (strncmp(key, EC_QUOTA_PREFIX, strlen(EC_QUOTA_PREFIX)) == 0) ||
@@ -225,15 +229,9 @@ int32_t ec_dict_list(data_t ** list, int32_t * count, ec_cbk_data_t * cbk,
 
         dict = (which == EC_COMBINE_XDATA) ? ans->xdata : ans->dict;
         list[i] = dict_get(dict, key);
-        if (list[i] == NULL)
-        {
-            gf_log(cbk->fop->xl->name, GF_LOG_ERROR, "Unexpected missing "
-                                                     "dictionary entry");
-
-            return 0;
+        if (list[i] != NULL) {
+            i++;
         }
-
-        i++;
     }
 
     *count = i;
@@ -471,11 +469,6 @@ int32_t ec_dict_data_max32(ec_cbk_data_t *cbk, int32_t which, char *key)
         return -1;
     }
 
-    if (num <= 1)
-    {
-        return 0;
-    }
-
     max = data_to_uint32(data[0]);
     for (i = 1; i < num; i++)
     {
@@ -507,10 +500,6 @@ int32_t ec_dict_data_max64(ec_cbk_data_t *cbk, int32_t which, char *key)
         return -1;
     }
 
-    if (num <= 1) {
-        return 0;
-    }
-
     max = data_to_uint64(data[0]);
     for (i = 1; i < num; i++) {
         tmp = data_to_uint64(data[i]);
@@ -630,6 +619,10 @@ int32_t ec_dict_data_combine(dict_t * dict, char * key, data_t * value,
     {
         return ec_dict_data_max32(data->cbk, data->which, key);
     }
+    if ((strcmp(key, GLUSTERFS_INODELK_COUNT) == 0) ||
+        (strcmp(key, GLUSTERFS_ENTRYLK_COUNT) == 0)) {
+        return ec_dict_data_max32(data->cbk, data->which, key);
+    }
 
     if (strcmp(key, QUOTA_SIZE_KEY) == 0) {
         return ec_dict_data_quota(data->cbk, data->which, key);
@@ -831,6 +824,8 @@ void ec_combine (ec_cbk_data_t *newcbk, ec_combine_f combine)
 
     LOCK(&fop->lock);
 
+    fop->received |= newcbk->mask;
+
     item = fop->cbk_list.prev;
     list_for_each_entry(cbk, &fop->cbk_list, list)
     {
@@ -868,7 +863,9 @@ void ec_combine (ec_cbk_data_t *newcbk, ec_combine_f combine)
     }
 
     cbk = list_entry(fop->cbk_list.next, ec_cbk_data_t, list);
-    needed = fop->minimum - cbk->count - fop->winds + 1;
+    if ((fop->mask ^ fop->remaining) == fop->received) {
+        needed = fop->minimum - cbk->count;
+    }
 
     UNLOCK(&fop->lock);