clickhouse与mysql查询速度对比

数据准备工作：

    MySQL创建准备对比的表Bigtable并建立索引

CREATE TABLE Bigtable (
consumption_id VARCHAR(255),
member_id INT(11),
fans_id INT(11),
bill_date DATE,
money VARCHAR(255),
people_num INT(8),
dish_name VARCHAR(255),
created_org INT(8),
open_id VARCHAR(255),
subscribed_time DATETIME,
unsubscribed_time DATETIME,
source_type INT(8),
sns_type VARCHAR(255),
is_subscribed VARCHAR(255)
)ENGINE=INNODB DEFAULT CHARSET=utf8;

CREATE INDEX indexbigtable ON bigtable(consumption_id,member_id,bill_date,created_org,source_type,sns_type,is_subscribed);

   将数据插入Bigtable，共308万行。

INSERT INTO bigtable SELECT a.consumption_id,a.member_id,b.fans_id,DATE(a.bill_date),a.need_pay_amount,a.people_num,a.dish_name,a.created_org,b.open_id,b.subscribed_time,
b.unsubscribed_time,b.source_type,b.sns_type,b.is_subscribed
FROM (
SELECT conn.consumption_id,conn.member_id,dish.dish_name,conn.bill_date,conn.people_num,conn.need_pay_amount,conn.created_org
FROM   `pos_consumption` AS conn 
INNER JOIN pos_dining_order AS porder ON conn.consumption_id = porder.consumption_id AND conn.created_org = porder.created_org
INNER JOIN pos_order_dish AS dish ON porder.order_id=dish.dining_order_id AND dish.created_org=porder.created_org
WHERE conn.`status_code`=1 
AND conn.`need_pay_amount` >0
) a
LEFT JOIN (SELECT fan.fans_id,fan.member_id,bind.open_id,bind.subscribed_time,bind.unsubscribed_time,fan.source_type,sns_type,is_subscribed   FROM rv_ol_fans AS fan 
INNER JOIN `0001790455_baseinfo`.`ol_fans_sns_bind` AS bind USE INDEX (idx_fans_id) ON fan.fans_id=bind.fans_id   WHERE bind.status_code=1)b ON a.member_id=b.member_id

   clickhouse建表bigtable，大小写一定要严格符合：

CREATE TABLE bigtable (\
consumption_id UInt64,\
member_id UInt64,\
fans_id UInt64,\
bill_date Date,\
money Float32,\
people_num UInt8,\
dish_name String,\
created_org UInt8,\
open_id String,\
subscribed_time DateTime,\
unsubscribed_time DateTime,\
source_type UInt8,\
sns_type UInt8,\
is_subscribed UInt8\
)ENGINE=MergeTree(bill_date,(consumption_id,created_org),8192)

mysql导出数据到csv：

SELECT consumption_id, 
member_id, fans_id, bill_date,
 money, people_num, 
 dish_name, created_org,open_id,subscribed_time,unsubscribed_time,
source_type,sns_type,is_subscribed
INTO OUTFILE '/var/lib/mysql-files/bigtable.csv'    
FIELDS TERMINATED BY ','  OPTIONALLY ENCLOSED BY '"'    
LINES TERMINATED BY '\n'   FROM 
bigtable; 

服务器之间的csv拷贝：

scp bigtable.csv root@192.168.x.xxx:/root/clickhouse-files

clickhouse导入csv：

cat /root/clickhouse-files/bigtable.csv | clickhouse-client --query="INSERT INTO test.bigtable FORMAT CSV"

数据量为308万行，每行14个列的测试：

• select * from bigtable

  clickhouse：67.62s   ；    mysql：0.002s

   clickhouse感觉是瞬间查出来，但是加载10000行数据刷屏用了很久很久，clickhouse里查询的列数一多，加载的时间就变长

• TOPN

  select* from bigtable order by dish_Name limit 1

  clickhouse：0.13s   ；    mysql：0.020s
  

• 统计数据有多少行：

select count(1) from bigtable

    clickhouse：0.015s   ；    mysql：1.33s

• 统计一共有多少个订单：

select count(1) from (select consumption_id,created_org from bigtable group by consumption_id,created_org)a;

    clickhouse：0.121s   ；    mysql：5.15s

• 来吃过的次数里各有多少人：82行
  

select sum_all,count(1) as countall from (select member_id,count(1) as sum_all from 
(select member_id,consumption_id,created_org from bigtable group by consumption_id,created_org) a group by member_id) a 
group by sum_all；

    clickhouse：0.166s   ；    mysql：5.50s

• 添加条件的查询：

SELECT COUNT(1) FROM bigtable 
WHERE member_id IN (1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,
20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,
40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,
60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,
76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,
95,96,97,98,99,100)
AND bill_date BETWEEN '2017-01-01' AND '2017-12-31'
AND fans_id >10

     clickhouse：0.027s   ；    mysql：2.05s

• 每个用户最常吃的菜品：8万行

SELECT member_id,dish_name,sum_all FROM (
SELECT member_id,dish_name,COUNT(1) AS sum_all FROM bigtable WHERE dish_name NOT IN('米饭','打包盒','茶位费') 
GROUP BY member_id,dish_name ORDER BY member_id,sum_all DESC) a GROUP BY member_id

clickhouse：11.90s ； mysql：9.88s

• 来个猛的，四重子查询：171万行

SELECT * FROM bigtable
WHERE member_id IN(SELECT member_id FROM bigtable WHERE member_id IN(SELECT member_id FROM bigtable WHERE source_type=1 AND member_id IN (SELECT member_id FROM bigtable WHERE source_type=1
AND member_id IN (SELECT member_id FROM bigtable WHERE fans_id !=0))
)  AND is_subscribed=1 )

clickhouse：65.284s ； mysql：

Mysql不行了，查了30分钟还没结果；clickhouse同样是加载行数用了很久

• 再来一个二重子查询试试：

SELECT * FROM bigtable\
WHERE member_id IN(SELECT member_id FROM bigtable WHERE member_id IN(SELECT member_id FROM bigtable WHERE source_type=1 )\
 AND is_subscribed=1 )

clickhouse：63.216s ； mysql：

没想到两个子查询的mysql在30分钟内也出不来结果

那么这部分对比查询时间的结果显而易见了：

    对于简单查询来说，查询列数多的话mysql有优势，查询列数少的话clickhouse领先。

    对于复杂查询来说，clickhouse占有显著优势

另外，展示行数的多少会影响clickhouse的查询时间，不知道是不是因为使用linux的原因

数据量为1亿两千四百万行，每行62个列的测试：

• 同样，先是select一下全表

select * from pdish_test

clickhouse：276s ； mysql：0.036s

• 统计行数：

select count(1) from pdish_test

clickhouse：0.044s ； mysql：32.168s

• 每个菜品点了多少次，共收入多少钱：377行

SELECT dish_name,COUNT(1) AS cc,SUM(dish_sale_amount) AS c FROM pdish_test GROUP BY dish_name ORDER BY cc DESC

clickhouse：3.55s ； mysql：8min6s

• 简单的条件查询：

 SELECT COUNT(1) FROM pdish_test WHERE dish_sale_amount>10 AND created_on >'2017-01-01'

clickhouse：0.448s ； mysql：

依然是30分钟过去了，mysql还是没有反应

看来对非常大的数据量来说，Mysql已经跑不动了

关联表查询，三个表分别为近100万，13万，13万：

    三个表相互关联拼接的查询： 8万行

clickhouse：3.65s ； mysql：4min46s

在几万十几万行的数据里，clickhouse的速度也是要显著快于mysql