Remove usage of row count (as it trashes _all_ explain & table status)

Use order by


mysql-test/r/ndb_index_ordered.result:
  Use order by
mysql-test/t/ndb_index_ordered.test:
  Use order by
sql/ha_ndbcluster.cc:
  Remove usage of row count (as it trashes _all_ explain & table status)

mysql-test/r/ndb_index_ordered.result

@@ -219,10 +219,10 @@ c int unsigned,
 KEY bc(b,c)
 ) engine = ndb;
 insert into t1 values(1,1,1),(2,NULL,2),(3,NULL,NULL),(4,4,NULL);
-select * from t1 use index (bc) where b IS NULL;
+select * from t1 use index (bc) where b IS NULL order by a;
 a	b	c
-3	NULL	NULL
 2	NULL	2
+3	NULL	NULL
 select * from t1 use index (bc)order by a;
 a	b	c
 1	1	1

mysql-test/t/ndb_index_ordered.test

@@ -130,7 +130,7 @@ CREATE TABLE t1 (
 ) engine = ndb;
 
 insert into t1 values(1,1,1),(2,NULL,2),(3,NULL,NULL),(4,4,NULL);
-select * from t1 use index (bc) where b IS NULL;
+select * from t1 use index (bc) where b IS NULL order by a;
 
 select * from t1 use index (bc)order by a;
 select * from t1 use index (bc) order by a;

sql/ha_ndbcluster.cc

@@ -557,7 +557,7 @@ int ha_ndbcluster::get_metadata(const char *path)
   // All checks OK, lets use the table
   m_table= (void*)tab;
   Uint64 rows;
-  if(ndb_get_table_statistics(m_ndb, m_tabname, &rows, 0) == 0){
+  if(false && ndb_get_table_statistics(m_ndb, m_tabname, &rows, 0) == 0){
     records= rows;
   }