利用Django框架中select_related和prefetch_related函數(shù)對(duì)數(shù)據(jù)庫(kù)查詢優(yōu)化

更新時(shí)間：2015年04月01日 15:38:49 作者：CuGBabyBeaR

這篇文章主要介紹了利用Python的Django框架中select_related和prefetch_related函數(shù)對(duì)數(shù)據(jù)庫(kù)查詢的優(yōu)化的一個(gè)實(shí)踐例子,展示如何在實(shí)際中利用這兩個(gè)函數(shù)減少對(duì)數(shù)據(jù)庫(kù)的查詢次數(shù),需要的朋友可以參考下

實(shí)例的背景說明

假定一個(gè)個(gè)人信息系統(tǒng)，需要記錄系統(tǒng)中各個(gè)人的故鄉(xiāng)、居住地、以及到過的城市。數(shù)據(jù)庫(kù)設(shè)計(jì)如下：

201541150650059.jpg (591×250)

Models.py 內(nèi)容如下：

from django.db import models
 
class Province(models.Model):
 name = models.CharField(max_length=10)
 def __unicode__(self):
  return self.name
 
class City(models.Model):
 name = models.CharField(max_length=5)
 province = models.ForeignKey(Province)
 def __unicode__(self):
  return self.name
 
class Person(models.Model):
 firstname = models.CharField(max_length=10)
 lastname = models.CharField(max_length=10)
 visitation = models.ManyToManyField(City, related_name = "visitor")
 hometown = models.ForeignKey(City, related_name = "birth")
 living  = models.ForeignKey(City, related_name = "citizen")
 def __unicode__(self):
  return self.firstname + self.lastname

注1：創(chuàng)建的app名為“QSOptimize”

注2：為了簡(jiǎn)化起見，`qsoptimize_province` 表中只有2條數(shù)據(jù)：湖北省和廣東省，`qsoptimize_city`表中只有三條數(shù)據(jù)：武漢市、十堰市和廣州市

如果我們想要獲得所有家鄉(xiāng)是湖北的人，最無(wú)腦的做法是先獲得湖北省，再獲得湖北的所有城市，最后獲得故鄉(xiāng)是這個(gè)城市的人。就像這樣：

>>> hb = Province.objects.get(name__iexact=u"湖北省")
>>> people = []
>>> for city in hb.city_set.all():
... people.extend(city.birth.all())
...

顯然這不是一個(gè)明智的選擇，因?yàn)檫@樣做會(huì)導(dǎo)致1+（湖北省城市數(shù)）次SQL查詢。反正是個(gè)反例，導(dǎo)致的查詢和獲得掉結(jié)果就不列出來(lái)了。
prefetch_related() 或許是一個(gè)好的解決方法，讓我們來(lái)看看。

>>> hb = Province.objects.prefetch_related("city_set__birth").objects.get(name__iexact=u"湖北省")
>>> people = []
>>> for city in hb.city_set.all():
... people.extend(city.birth.all())
...

因?yàn)槭且粋€(gè)深度為2的prefetch，所以會(huì)導(dǎo)致3次SQL查詢：

SELECT `QSOptimize_province`.`id`, `QSOptimize_province`.`name`
FROM `QSOptimize_province`
WHERE `QSOptimize_province`.`name` LIKE '湖北省' ;
 
SELECT `QSOptimize_city`.`id`, `QSOptimize_city`.`name`, `QSOptimize_city`.`province_id`
FROM `QSOptimize_city`
WHERE `QSOptimize_city`.`province_id` IN (1);
 
SELECT `QSOptimize_person`.`id`, `QSOptimize_person`.`firstname`, `QSOptimize_person`.`lastname`,
`QSOptimize_person`.`hometown_id`, `QSOptimize_person`.`living_id`
FROM `QSOptimize_person`
WHERE `QSOptimize_person`.`hometown_id` IN (1, 3);

嗯…看上去不錯(cuò)，但是3次查詢么？倒過來(lái)查詢可能會(huì)更簡(jiǎn)單？

>>> people = list(Person.objects.select_related("hometown__province").filter(hometown__province__name__iexact=u"湖北省"))
 
SELECT `QSOptimize_person`.`id`, `QSOptimize_person`.`firstname`, `QSOptimize_person`.`lastname`,
`QSOptimize_person`.`hometown_id`, `QSOptimize_person`.`living_id`, `QSOptimize_city`.`id`,
`QSOptimize_city`.`name`, `QSOptimize_city`.`province_id`, `QSOptimize_province`.`id`, `QSOptimize_province`.`name`
FROM `QSOptimize_person`
INNER JOIN `QSOptimize_city` ON (`QSOptimize_person`.`hometown_id` = `QSOptimize_city`.`id`)
INNER JOIN `QSOptimize_province` ON (`QSOptimize_city`.`province_id` = `QSOptimize_province`.`id`)
WHERE `QSOptimize_province`.`name` LIKE '湖北省';
 
+----+-----------+----------+-------------+-----------+----+--------+-------------+----+--------+
| id | firstname | lastname | hometown_id | living_id | id | name | province_id | id | name |
+----+-----------+----------+-------------+-----------+----+--------+-------------+----+--------+
| 1 | 張  | 三  |   3 |   1 | 3 | 十堰市 |   1 | 1 | 湖北省 |
| 2 | 李  | 四  |   1 |   3 | 1 | 武漢市 |   1 | 1 | 湖北省 |
| 3 | 王  | 麻子  |   3 |   2 | 3 | 十堰市 |   1 | 1 | 湖北省 |
+----+-----------+----------+-------------+-----------+----+--------+-------------+----+--------+
3 rows in set (0.00 sec)

完全沒問題。不僅SQL查詢的數(shù)量減少了，python程序上也精簡(jiǎn)了。
select_related()的效率要高于prefetch_related()。因此，最好在能用select_related()的地方盡量使用它，也就是說，對(duì)于ForeignKey字段，避免使用prefetch_related()。
聯(lián)用
對(duì)于同一個(gè)QuerySet，你可以同時(shí)使用這兩個(gè)函數(shù)。
在我們一直使用的例子上加一個(gè)model：Order （訂單）

class Order(models.Model):
 customer = models.ForeignKey(Person)
 orderinfo = models.CharField(max_length=50)
 time  = models.DateTimeField(auto_now_add = True)
 def __unicode__(self):
  return self.orderinfo

如果我們拿到了一個(gè)訂單的id 我們要知道這個(gè)訂單的客戶去過的省份。因?yàn)橛蠱anyToManyField顯然必須要用prefetch_related()。如果只用prefetch_related()會(huì)怎樣呢？

>>> plist = Order.objects.prefetch_related('customer__visitation__province').get(id=1)
>>> for city in plist.customer.visitation.all():
... print city.province.name
...

顯然，關(guān)系到了4個(gè)表：Order、Person、City、Province，根據(jù)prefetch_related()的特性就得有4次SQL查詢

SELECT `QSOptimize_order`.`id`, `QSOptimize_order`.`customer_id`, `QSOptimize_order`.`orderinfo`, `QSOptimize_order`.`time`
FROM `QSOptimize_order`
WHERE `QSOptimize_order`.`id` = 1 ;
 
SELECT `QSOptimize_person`.`id`, `QSOptimize_person`.`firstname`, `QSOptimize_person`.`lastname`, `QSOptimize_person`.`hometown_id`, `QSOptimize_person`.`living_id`
FROM `QSOptimize_person`
WHERE `QSOptimize_person`.`id` IN (1);
 
SELECT (`QSOptimize_person_visitation`.`person_id`) AS `_prefetch_related_val`, `QSOptimize_city`.`id`,
`QSOptimize_city`.`name`, `QSOptimize_city`.`province_id`
FROM `QSOptimize_city`
INNER JOIN `QSOptimize_person_visitation` ON (`QSOptimize_city`.`id` = `QSOptimize_person_visitation`.`city_id`)
WHERE `QSOptimize_person_visitation`.`person_id` IN (1);
 
SELECT `QSOptimize_province`.`id`, `QSOptimize_province`.`name`
FROM `QSOptimize_province`
WHERE `QSOptimize_province`.`id` IN (1, 2);

+----+-------------+---------------+---------------------+
| id | customer_id | orderinfo  | time    |
+----+-------------+---------------+---------------------+
| 1 |   1 | Info of Order | 2014-08-10 17:05:48 |
+----+-------------+---------------+---------------------+
1 row in set (0.00 sec)
 
+----+-----------+----------+-------------+-----------+
| id | firstname | lastname | hometown_id | living_id |
+----+-----------+----------+-------------+-----------+
| 1 | 張  | 三  |   3 |   1 |
+----+-----------+----------+-------------+-----------+
1 row in set (0.00 sec)
 
+-----------------------+----+--------+-------------+
| _prefetch_related_val | id | name | province_id |
+-----------------------+----+--------+-------------+
|      1 | 1 | 武漢市 |   1 |
|      1 | 2 | 廣州市 |   2 |
|      1 | 3 | 十堰市 |   1 |
+-----------------------+----+--------+-------------+
3 rows in set (0.00 sec)
 
+----+--------+
| id | name |
+----+--------+
| 1 | 湖北省 |
| 2 | 廣東省 |
+----+--------+
2 rows in set (0.00 sec)

更好的辦法是先調(diào)用一次select_related()再調(diào)用prefetch_related()，最后再select_related()后面的表

>>> plist = Order.objects.select_related('customer').prefetch_related('customer__visitation__province').get(id=1)
>>> for city in plist.customer.visitation.all():
... print city.province.name
...

這樣只會(huì)有3次SQL查詢，Django會(huì)先做select_related，之后prefetch_related的時(shí)候會(huì)利用之前緩存的數(shù)據(jù)，從而避免了1次額外的SQL查詢：

SELECT `QSOptimize_order`.`id`, `QSOptimize_order`.`customer_id`, `QSOptimize_order`.`orderinfo`, 
`QSOptimize_order`.`time`, `QSOptimize_person`.`id`, `QSOptimize_person`.`firstname`, 
`QSOptimize_person`.`lastname`, `QSOptimize_person`.`hometown_id`, `QSOptimize_person`.`living_id` 
FROM `QSOptimize_order` 
INNER JOIN `QSOptimize_person` ON (`QSOptimize_order`.`customer_id` = `QSOptimize_person`.`id`) 
WHERE `QSOptimize_order`.`id` = 1 ;
 
SELECT (`QSOptimize_person_visitation`.`person_id`) AS `_prefetch_related_val`, `QSOptimize_city`.`id`, 
`QSOptimize_city`.`name`, `QSOptimize_city`.`province_id` 
FROM `QSOptimize_city` 
INNER JOIN `QSOptimize_person_visitation` ON (`QSOptimize_city`.`id` = `QSOptimize_person_visitation`.`city_id`) 
WHERE `QSOptimize_person_visitation`.`person_id` IN (1);
 
SELECT `QSOptimize_province`.`id`, `QSOptimize_province`.`name` 
FROM `QSOptimize_province` 
WHERE `QSOptimize_province`.`id` IN (1, 2);
 
+----+-------------+---------------+---------------------+----+-----------+----------+-------------+-----------+
| id | customer_id | orderinfo  | time    | id | firstname | lastname | hometown_id | living_id |
+----+-------------+---------------+---------------------+----+-----------+----------+-------------+-----------+
| 1 |   1 | Info of Order | 2014-08-10 17:05:48 | 1 | 張  | 三  |   3 |   1 |
+----+-------------+---------------+---------------------+----+-----------+----------+-------------+-----------+
1 row in set (0.00 sec)
 
+-----------------------+----+--------+-------------+
| _prefetch_related_val | id | name | province_id |
+-----------------------+----+--------+-------------+
|      1 | 1 | 武漢市 |   1 |
|      1 | 2 | 廣州市 |   2 |
|      1 | 3 | 十堰市 |   1 |
+-----------------------+----+--------+-------------+
3 rows in set (0.00 sec)
 
+----+--------+
| id | name |
+----+--------+
| 1 | 湖北省 |
| 2 | 廣東省 |
+----+--------+
2 rows in set (0.00 sec)

值得注意的是，可以在調(diào)用prefetch_related之前調(diào)用select_related，并且Django會(huì)按照你想的去做：先select_related，然后利用緩存到的數(shù)據(jù)prefetch_related。然而一旦prefetch_related已經(jīng)調(diào)用，select_related將不起作用。

小結(jié)

因?yàn)閟elect_related()總是在單次SQL查詢中解決問題，而prefetch_related()會(huì)對(duì)每個(gè)相關(guān)表進(jìn)行SQL查詢，因此select_related()的效率通常比后者高。
鑒于第一條，盡可能的用select_related()解決問題。只有在select_related()不能解決問題的時(shí)候再去想prefetch_related()。
你可以在一個(gè)QuerySet中同時(shí)使用select_related()和prefetch_related()，從而減少SQL查詢的次數(shù)。
只有prefetch_related()之前的select_related()是有效的，之后的將會(huì)被無(wú)視掉。

您可能感興趣的文章: