branches/zip:

The cardinality of every index (the number of different key values) is
calculated when the table is opened, at SHOW TABLE STATUS,
ANALYZE TABLE and on other circumstances (like when the table has
changed too much). Note that if the mysql client is running with the
auto-rehash setting turned on (default) this causes all tables to be
opened when it starts.

Previously InnoDB sampled 8 random pages from the index to get an
estimate of the cardinality. Now the number of sampled pages can be
changed via the global parameter innodb_stats_sample_pages which can
be tuned at runtime. The default value for this parameter is 8.

If the value of this parameter is changed, there may be serious problems:

- small values (say, 1) can cause an error in table stats;
- values much larger than 8 (say, 100), can cause a big slowdown in
  table opening time, SHOW TABLE status, etc.
- query plans may be different from the old ones.

Approved by:	Heikki

btr/btr0cur.c

@@ -56,10 +56,6 @@ can be released by page reorganize, then it is reorganized */
 
 #define BTR_CUR_PAGE_REORGANIZE_LIMIT	(UNIV_PAGE_SIZE / 32)
 
-/* When estimating number of different key values in an index, sample
-this many index pages */
-#define BTR_KEY_VAL_ESTIMATE_N_PAGES	8
-
 /* The structure of a BLOB part header */
 /*--------------------------------------*/
 #define BTR_BLOB_HDR_PART_LEN		0	/* BLOB part len on this
@@ -3175,7 +3171,7 @@ btr_estimate_number_of_different_key_vals(
 
 	/* We sample some pages in the index to get an estimate */
 
-	for (i = 0; i < BTR_KEY_VAL_ESTIMATE_N_PAGES; i++) {
+	for (i = 0; i < srv_stats_sample_pages; i++) {
 		rec_t*	supremum;
 		mtr_start(&mtr);
 
@@ -3264,7 +3260,7 @@ btr_estimate_number_of_different_key_vals(
 	}
 
 	/* If we saw k borders between different key values on
-	BTR_KEY_VAL_ESTIMATE_N_PAGES leaf pages, we can estimate how many
+	srv_stats_sample_pages leaf pages, we can estimate how many
 	there will be in index->stat_n_leaf_pages */
 
 	/* We must take into account that our sample actually represents
@@ -3275,26 +3271,26 @@ btr_estimate_number_of_different_key_vals(
 		index->stat_n_diff_key_vals[j]
 			= ((n_diff[j]
 			    * (ib_int64_t)index->stat_n_leaf_pages
-			    + BTR_KEY_VAL_ESTIMATE_N_PAGES - 1
+			    + srv_stats_sample_pages - 1
 			    + total_external_size
 			    + not_empty_flag)
-			   / (BTR_KEY_VAL_ESTIMATE_N_PAGES
+			   / (srv_stats_sample_pages
 			      + total_external_size));
 
 		/* If the tree is small, smaller than
-		10 * BTR_KEY_VAL_ESTIMATE_N_PAGES + total_external_size, then
+		10 * srv_stats_sample_pages + total_external_size, then
 		the above estimate is ok. For bigger trees it is common that we
 		do not see any borders between key values in the few pages
-		we pick. But still there may be BTR_KEY_VAL_ESTIMATE_N_PAGES
+		we pick. But still there may be srv_stats_sample_pages
 		different key values, or even more. Let us try to approximate
 		that: */
 
 		add_on = index->stat_n_leaf_pages
-			/ (10 * (BTR_KEY_VAL_ESTIMATE_N_PAGES
+			/ (10 * (srv_stats_sample_pages
 				 + total_external_size));
 
-		if (add_on > BTR_KEY_VAL_ESTIMATE_N_PAGES) {
-			add_on = BTR_KEY_VAL_ESTIMATE_N_PAGES;
+		if (add_on > srv_stats_sample_pages) {
+			add_on = srv_stats_sample_pages;
 		}
 
 		index->stat_n_diff_key_vals[j] += add_on;

handler/ha_innodb.cc

@@ -9243,6 +9243,11 @@ static MYSQL_SYSVAR_BOOL(stats_on_metadata, innobase_stats_on_metadata,
   "Enable statistics gathering for metadata commands such as SHOW TABLE STATUS (on by default)",
   NULL, NULL, TRUE);
 
+static MYSQL_SYSVAR_ULONGLONG(stats_sample_pages, srv_stats_sample_pages,
+  PLUGIN_VAR_RQCMDARG,
+  "The number of index pages to sample when calculating statistics (default 8)",
+  NULL, NULL, 8, 1, ~0ULL, 0);
+
 static MYSQL_SYSVAR_BOOL(adaptive_hash_index, innobase_adaptive_hash_index,
   PLUGIN_VAR_OPCMDARG | PLUGIN_VAR_READONLY,
   "Enable InnoDB adaptive hash index (enabled by default).  "
@@ -9390,6 +9395,7 @@ static struct st_mysql_sys_var* innobase_system_variables[]= {
   MYSQL_SYSVAR(open_files),
   MYSQL_SYSVAR(rollback_on_timeout),
   MYSQL_SYSVAR(stats_on_metadata),
+  MYSQL_SYSVAR(stats_sample_pages),
   MYSQL_SYSVAR(adaptive_hash_index),
   MYSQL_SYSVAR(replication_delay),
   MYSQL_SYSVAR(status_file),

include/srv0srv.h

@@ -136,6 +136,8 @@ extern ibool	srv_innodb_status;
 
 extern ibool	srv_stats_on_metadata;
 
+extern unsigned long long	srv_stats_sample_pages;
+
 extern ibool	srv_use_doublewrite_buf;
 extern ibool	srv_use_checksums;
 

srv/srv0srv.c

@@ -303,6 +303,10 @@ UNIV_INTERN ibool	srv_innodb_status	= FALSE;
 
 UNIV_INTERN ibool	srv_stats_on_metadata	= TRUE;
 
+/* When estimating number of different key values in an index, sample
+this many index pages */
+UNIV_INTERN unsigned long long	srv_stats_sample_pages = 8;
+
 UNIV_INTERN ibool	srv_use_doublewrite_buf	= TRUE;
 UNIV_INTERN ibool	srv_use_checksums = TRUE;