2022-08-01 mysql/stoonedb慢SQL-Q18分析

摘要:

分析慢SQL-Q18出现的原因

DDL分析:

SQL:


select
	c_name,
	c_custkey,
	o_orderkey,
	o_orderdate,
	o_totalprice,
	sum(l_quantity)
from
	customer,
	orders,
	lineitem
where
	o_orderkey in (
	select
		l_orderkey
	from
		lineitem
	group by
		l_orderkey
	having
		sum(l_quantity) > 300 )
	and c_custkey = o_custkey
	and o_orderkey = l_orderkey
group by
	c_name,
	c_custkey,
	o_orderkey,
	o_orderdate,
	o_totalprice
order by
	o_totalprice desc,
	o_orderdate
limit 100;

innodb的explain分析:


mysql> explain select
    -> c_name,
    -> c_custkey,
    -> o_orderkey,
    -> o_orderdate,
    -> o_totalprice,
    -> sum(l_quantity)
    -> from
    -> customer,
    -> orders,
    -> lineitem
    -> where
    -> o_orderkey in (
    -> select
    -> 
Display all 1017 possibilities? (y or n) 
    -> l_orderkey
    -> from
    -> 
Display all 1017 possibilities? (y or n) 
    -> lineitem
    -> group by
    -> 
Display all 1017 possibilities? (y or n) 
    -> l_orderkey
    -> having
    -> 
Display all 1017 possibilities? (y or n) 
    -> sum(l_quantity) > 300 )
    -> and c_custkey = o_custkey
    -> and o_orderkey = l_orderkey
    -> group by
    -> c_name,
    -> c_custkey,
    -> o_orderkey,
    -> o_orderdate,
    -> o_totalprice
    -> order by
    -> o_totalprice desc,
    -> o_orderdate
    -> limit 100\G
*************************** 1. row ***************************
           id: 1
  select_type: PRIMARY
        table: orders
   partitions: NULL
         type: ALL
possible_keys: PRIMARY
          key: NULL
      key_len: NULL
          ref: NULL
         rows: 13754753
     filtered: 100.00
        Extra: Using where; Using temporary; Using filesort
*************************** 2. row ***************************
           id: 1
  select_type: PRIMARY
        table: customer
   partitions: NULL
         type: eq_ref
possible_keys: PRIMARY
          key: PRIMARY
      key_len: 4
          ref: tpch.orders.o_custkey
         rows: 1
     filtered: 100.00
        Extra: NULL
*************************** 3. row ***************************
           id: 1
  select_type: PRIMARY
        table: lineitem
   partitions: NULL
         type: ref
possible_keys: PRIMARY
          key: PRIMARY
      key_len: 4
          ref: tpch.orders.o_orderkey
         rows: 4
     filtered: 100.00
        Extra: NULL
*************************** 4. row ***************************
           id: 2
  select_type: SUBQUERY
        table: lineitem
   partitions: NULL
         type: index
possible_keys: PRIMARY
          key: PRIMARY
      key_len: 8
          ref: NULL
         rows: 31751480
     filtered: 100.00
        Extra: NULL
4 rows in set, 1 warning (0.01 sec)

select_type	table	partitions	type	possible_keys	key	key_len	ref	rows	filtered	Extra
PRIMARY	orders		ALL	PRIMARY				13754753	100	Using where; Using temporary; Using filesort
PRIMARY	customer		eq_ref	PRIMARY	PRIMARY	4	tpch.orders.o_custkey	1	100
PRIMARY	lineitem		ref	PRIMARY	PRIMARY	4	tpch.orders.o_orderkey	4	100
SUBQUERY	lineitem		index	PRIMARY	PRIMARY	8		31751480	100

stonedb的explain分析:


mysql> explain select
    -> c_name,
    -> c_custkey,
    -> o_orderkey,
    -> o_orderdate,
    -> o_totalprice,
    -> sum(l_quantity)
    -> from
    -> customer,
    -> orders,
    -> lineitem
    -> where
    -> o_orderkey in (
    -> select
    -> l_orderkey
    -> from
    -> lineitem
    -> group by
    -> l_orderkey
    -> having
    -> sum(l_quantity) > 300 )
    -> and c_custkey = o_custkey
    -> and o_orderkey = l_orderkey
    -> group by
    -> c_name,
    -> c_custkey,
    -> o_orderkey,
    -> o_orderdate,
    -> o_totalprice
    -> order by
    -> o_totalprice desc,
    -> o_orderdate
    -> limit 100\G
*************************** 1. row ***************************
           id: 1
  select_type: PRIMARY
        table: lineitem
   partitions: NULL
         type: ALL
possible_keys: PRIMARY
          key: NULL
      key_len: NULL
          ref: NULL
         rows: 59986052
     filtered: 100.00
        Extra: Using where with pushed condition <in_optimizer>(`tpch`.`lineitem`.`l_orderkey`,`tpch`.`lineitem`.`l_orderkey` in ( <materialize> (/* select#2 */ select `tpch`.`lineitem`.`l_orderkey` from `tpch`.`lineitem` group by `tpch`.`lineitem`.`l_orderkey` having (sum(`tpch`.`lineitem`.`l_quantity`) > 300) ), <primary_index_lookup>(`tpch`.`lineitem`.`l_orderkey` in <temporary table> on <auto_key> where ((`tpch`.`lineitem`.`l_orderkey` = `materialized-subquery`.`l_orderkey`)))))(t0) Pckrows: 916, susp. 916 (0 empty 0 full). Conditions: 1; Using temporary; Using filesort
*************************** 2. row ***************************
           id: 1
  select_type: PRIMARY
        table: orders
   partitions: NULL
         type: eq_ref
possible_keys: PRIMARY
          key: PRIMARY
      key_len: 4
          ref: tpch.lineitem.l_orderkey
         rows: 1
     filtered: 100.00
        Extra: NULL
*************************** 3. row ***************************
           id: 1
  select_type: PRIMARY
        table: customer
   partitions: NULL
         type: eq_ref
possible_keys: PRIMARY
          key: PRIMARY
      key_len: 4
          ref: tpch.orders.o_custkey
         rows: 1
     filtered: 100.00
        Extra: NULL
*************************** 4. row ***************************
           id: 2
  select_type: SUBQUERY
        table: lineitem
   partitions: NULL
         type: ALL
possible_keys: PRIMARY
          key: NULL
      key_len: NULL
          ref: NULL
         rows: 59986052
     filtered: 100.00
        Extra: Using temporary; Using filesort
4 rows in set, 1 warning (54.11 sec)

select_type

table

partitions

type

possible_keys

key

key_len

ref

rows

filtered

Extra

PRIMARY

lineitem

ALL

PRIMARY

59986052

100

Using where with pushed condition (`tpch`.`lineitem`.`l_orderkey`,`tpch`.`lineitem`.`l_orderkey` in ( (/* select#2 */ select `tpch`.`lineitem`.`l_orderkey` from `tpch`.`lineitem` group by `tpch`.`lineitem`.`l_orderkey` having (sum(`tpch`.`lineitem`.`l_quantity`) > 300) ), (`tpch`.`lineitem`.`l_orderkey` in on where ((`tpch`.`lineitem`.`l_orderkey` = `materialized-subquery`.`l_orderkey`)))))(t0) Pckrows: 916, susp. 916 (0 empty 0 full). Conditions: 1; Using temporary; Using filesort

PRIMARY

orders

eq_ref

PRIMARY

tpch.lineitem.l_orderkey

100

PRIMARY

customer

eq_ref

PRIMARY

tpch.orders.o_custkey

100

SUBQUERY

lineitem

ALL

PRIMARY

59986052

100

Using temporary; Using filesort

执行分析:

热力图:

hash join数据统计:


[2022-08-01 19:27:48.120105] [1369856] End traversed 8380864/8380864 rows.
[2022-08-01 19:28:28.177049] [1369856] Begin match dim of 32986052 rows, spliting into 64 threads with packs type.
[2022-08-01 19:29:07.932297] [1369856] End match dim. Produced tuples:32986052/32986052
[2022-08-01 19:29:07.985311] [1369856] Tuples after inner join 1-2 [hash]: 32986052 	[111]
[2022-08-01 19:29:08.286386] [1369856] GroupTable begin, initialized for up to 32986052 groups, 53+31 bytes (3171 MB)
[2022-08-01 19:29:08.529321] [1369856] Aggregating: 32986052 tuples left.
[2022-08-01 19:30:40.828429] [1369856] Group/Aggregate end. Begin generating output.
[2022-08-01 19:30:40.828502] [1369856] Output rows: 8249593, output table row limit: 8249593
[2022-08-01 19:31:06.163156] [1369856] Generating output end. Aggregated (8249593 group). Omitted packrows: 0 + 0 partially, out of 26622 total.
[2022-08-01 19:31:08.930503] [1369856] Heap Sort begin, initialized for 8249593 rows, 16+0+3 bytes each.
[2022-08-01 19:32:21.876613] [1369856] Sorted end, rows retrieved.

相关阅读:
Zero-Shot Learning by Harnessing Adversarial Samples 理论 & 代码解读
 自学Python01-创建文件写入内容
 【活动系列】那些年写的比较愚蠢的代码
 The Missing Semester of Your CS Education(计算机教育中缺失的一课)
python自动化第一篇—— 带图文的execl的自动化合并
 没有网络怎么配置传奇SF登陆器自动读取列表
 配置本地Maven仓库——IDEA配置本地Maven源
 Inno Setup安装中文语言
 01-微服务探讨(摘)
方法练习（二）
原文地址：https://blog.csdn.net/adofsauron/article/details/126098067