通过profile命令来查看当前最主要的耗费时间的步骤。
mysql> select count(1) from t1;
+----------+
| count(1) |
+----------+
| 1 |
+----------+
1 row in set (0.11 sec)
mysql> show profiles;
+----------+------------+-------------------------+
| Query_ID | Duration | Query |
+----------+------------+-------------------------+
| 1 | 0.00123300 | show databases |
| 2 | 0.00016775 | SELECT DATABASE() |
| 3 | 0.00092900 | show databases |
| 4 | 0.00122325 | show tables |
| 5 | 0.00134250 | show tables |
| 6 | 0.11396400 | select count(1) from t1 |
+----------+------------+-------------------------+
6 rows in set, 1 warning (0.00 sec)
mysql> show profile for query 6;
+--------------------------------+----------+
| Status | Duration |
+--------------------------------+----------+
| starting | 0.000089 |
| Executing hook on transaction | 0.000013 |
| starting | 0.000011 |
| checking permissions | 0.000008 |
| Opening tables | 0.000040 |
| init | 0.000015 |
| System lock | 0.000015 |
| optimizing | 0.000007 |
| statistics | 0.000024 |
| preparing | 0.000029 |
| executing | 0.113622 |
| end | 0.000011 |
| query end | 0.000006 |
| waiting for handler commit | 0.000014 |
| closing tables | 0.000013 |
| freeing items | 0.000022 |
| cleaning up | 0.000027 |
+--------------------------------+----------+
17 rows in set, 1 warning (0.00 sec)mysql> select state,sum(duration) as total_r , round(100*sum(duration)/(select sum(duration) from information_schema.profiling where query_id=6),2) as pct_r,count(*) as calls,sum(duration)/count(*) as 'R/Call' from information_schema.profiling where query_id group by state order by total_r desc;
+--------------------------------+----------+--------+-------+--------------+
| state | total_r | pct_r | calls | R/Call |
+--------------------------------+----------+--------+-------+--------------+
| executing | 0.114058 | 100.08 | 6 | 0.0190096667 |
| Opening tables | 0.001226 | 1.08 | 6 | 0.0002043333 |
| checking permissions | 0.000852 | 0.75 | 56 | 0.0000152143 |
| starting | 0.000692 | 0.61 | 8 | 0.0000865000 |
| Creating tmp table | 0.000524 | 0.46 | 4 | 0.0001310000 |
| init | 0.000410 | 0.36 | 6 | 0.0000683333 |
| statistics | 0.000361 | 0.32 | 5 | 0.0000722000 |
| preparing | 0.000165 | 0.14 | 5 | 0.0000330000 |
| freeing items | 0.000138 | 0.12 | 7 | 0.0000197143 |
| cleaning up | 0.000108 | 0.09 | 7 | 0.0000154286 |
| optimizing | 0.000092 | 0.08 | 6 | 0.0000153333 |
| waiting for handler commit | 0.000084 | 0.07 | 9 | 0.0000093333 |
| System lock | 0.000079 | 0.07 | 5 | 0.0000158000 |
| closing tables | 0.000067 | 0.06 | 6 | 0.0000111667 |
| query end | 0.000039 | 0.03 | 6 | 0.0000065000 |
| end | 0.000035 | 0.03 | 6 | 0.0000058333 |
| removing tmp table | 0.000018 | 0.02 | 4 | 0.0000045000 |
| Executing hook on transaction | 0.000013 | 0.01 | 1 | 0.0000130000 |
+--------------------------------+----------+--------+-------+--------------+
18 rows in set, 2 warnings (0.00 sec)
profile还支持查看all、cpu、block io、content switch、page faults等明细类型，例如查看上述语句在CPU资源上消耗的时间：
show profile cpu for query 6；
