sampling analysiskakurin_sampling

file=open(...)
for l in file:
     if l.startswith('SAMPLE') and not l.startswith('SAMPLE root'):
         print(l)
         n = dict((p.split('=')
           for p in l.replace('SAMPLE ', '').replace(' REJECT', '').split()
           ))
         if 'REJECT' in l:
             n['rej']=True
         nodes.append(n)

max(int(n['size']) for n in nodes)
max((float(n['p']), serialize(n)) for n in nodes)
max((float(n['size']), serialize(n)) for n in nodes if n['height'] == '0')

4 files changed, 26 insertions, 36 deletions

diff --git a/mysql-test/main/profile.opt b/mysql-test/main/profile.opt
index da84f76d539..8fbae672382 100644
--- a/mysql-test/main/profile.opt
+++ b/mysql-test/main/profile.opt
@@ -1,2 +1,2 @@
---innodb_buffer_pool_size=8M
+--innodb_buffer_pool_size=1G
 
diff --git a/mysql-test/main/profile.test b/mysql-test/main/profile.test
index dcff6e480ae..b1c7bc21d9a 100644
--- a/mysql-test/main/profile.test
+++ b/mysql-test/main/profile.test
@@ -50,7 +50,7 @@ create or replace temporary table counters (
 ) engine=myisam;
 
 
---let $i= 500
+--let $i= 5
 --echo i=$i
 --let $method=bernoulli
 --let percent=1
@@ -67,36 +67,4 @@ create or replace temporary table counters (
 --enable_result_log
 select pk_value div 10000, sum(counter) from counters group by pk_value div 10000;
 
-create or replace temporary table counters (
-  pk_value int,
-  counter int,
-  primary key (pk_value)
-) engine=myisam;
-
-
---echo ==================================================================
---echo ==================================================================
---echo ==================================================================
---echo ==================================================================
---echo ==================================================================
---echo ==================================================================
---echo ==================================================================
---echo ==================================================================
-
---let $i= 500
---echo i=$i
---let $method=system
---let percent=1
---echo method=$method $percent%
---while($i)
-  eval
-INSERT INTO counters
-  SELECT _id, 0 FROM vehicle_reg tablesample $method($percent)
-ON DUPLICATE KEY UPDATE counter = counter + 1
-;
---dec $i
---}
---enable_query_log
---enable_result_log
-select pk_value div 10000, sum(counter) from counters group by pk_value div 10000;
 --}
diff --git a/storage/innobase/btr/btr0cur.cc b/storage/innobase/btr/btr0cur.cc
index 04ee06bf032..027ef598d50 100644
--- a/storage/innobase/btr/btr0cur.cc
+++ b/storage/innobase/btr/btr0cur.cc
@@ -2847,6 +2847,7 @@ btr_cur_open_at_index_side(
 	return err;
 }
 extern unsigned long long my_counter;
+ulint get_last_nth();
 /**********************************************************************//**
 Positions a cursor at a randomly chosen position within a B-tree.
 @return DB_SUCCESS if the index is available and we have put the cursor,
@@ -2952,8 +2953,10 @@ btr_cur_open_at_rnd_pos(
 	}
 
 	height = ULINT_UNDEFINED;
+        fprintf(stderr, "SAMPLE root page_id=%d\n", page_id.page_no());
 
 	for (;;) {
+		bool root= false;
 		page_t*		page;
 
 		ut_ad(n_blocks < BTR_MAX_LEVELS);
@@ -3012,6 +3015,7 @@ btr_cur_open_at_rnd_pos(
 			/* We are in the root node */
 
 			height = btr_page_get_level(page);
+                        root= true;
 		} else {
 			if(sim_uniform_dist) {
 				ulint n_recs = page_get_n_recs(block->page.frame);
@@ -3059,6 +3063,12 @@ btr_cur_open_at_rnd_pos(
 
                 page_cur_open_on_rnd_user_rec(block, page_cursor);
 
+                fprintf(stderr, "SAMPLE node=%x root=%d p=%lf height=%ld size=%d child_no=%ld ",
+                        page_id.page_no(), root,
+                        p,
+                        height,
+                        page_get_n_recs(block->page.frame),
+                        get_last_nth());
 		if (height == 0) {
 
 			break;
@@ -3145,6 +3155,8 @@ btr_cur_open_at_rnd_pos(
 		page_id.set_page_no(
 			btr_node_ptr_get_child_page_no(node_ptr, offsets));
 
+                fprintf(stderr, "child_page=%x\n", page_id.page_no());
+
 		n_blocks++;
 	}
 
@@ -3156,10 +3168,12 @@ btr_cur_open_at_rnd_pos(
 	// getting 0..1 pseudo random number from ut_rnd_gen()
 	// and exchange division by multiplication like
 	// (b / c) < a <=> b < (a * c)
-	if(sim_uniform_dist && (ut_rnd_gen() < (p * ~(uint32_t)0))) {
+	if(sim_uniform_dist && (ut_rnd_gen() > (p * ~(uint32_t)0))) {
 		err = DB_RECORD_NOT_FOUND;
                 my_counter++;
+                fprintf(stderr, "REJECT\n");
         }
+        else fprintf(stderr, "\n");
 
 	return err;
 }
diff --git a/storage/innobase/page/page0cur.cc b/storage/innobase/page/page0cur.cc
index 41ae2638f47..ea5bd0da007 100644
--- a/storage/innobase/page/page0cur.cc
+++ b/storage/innobase/page/page0cur.cc
@@ -764,6 +764,13 @@ up_rec_match:
 }
 #endif /* BTR_CUR_HASH_ADAPT */
 
+
+thread_local ulint last_nth;
+ulint get_last_nth()
+{
+  return last_nth;
+}
+
 /***********************************************************//**
 Positions a page cursor on a randomly chosen user record on a page. If there
 are no user records, sets the cursor on the infimum record. */
@@ -782,8 +789,9 @@ page_cur_open_on_rnd_user_rec(
 		return;
 	}
 
+        last_nth= ut_rnd_interval(n_recs) + 1;
 	cursor->rec = page_rec_get_nth(block->page.frame,
-				       ut_rnd_interval(n_recs) + 1);
+				       last_nth);
 }
 
 /**