Django数据库查询优化（一）

最新推荐文章于 2023-06-27 09:14:23 发布

pipipippipipipipip

最新推荐文章于 2023-06-27 09:14:23 发布

阅读量615

点赞数

文章标签： Django

本文链接：https://blog.csdn.net/weixin_43182689/article/details/95320116

版权

数据库查询优化（一）

收藏的一篇文章
下面内容会使用到的模型类，最近在看TheDjangoBook，这个模型类是里面的，稍微做了些改动

# models.py
from django.db import models

# Create your models here.


class Publisher(models.Model):
    name = models.CharField(max_length=30)
    address = models.CharField(max_length=50)
    city = models.CharField(max_length=60)
    state_province = models.CharField(max_length=30)
    # country = models.CharField(max_length=50)
    country = models.ForeignKey('Country')
    website = models.URLField()

    def __str__(self):
        return self.name


class Auther(models.Model):
    first_name = models.CharField(max_length=30)
    last_name = models.CharField(max_length=30)
    email = models.EmailField()


    def __str__(self):
        return self.first_name


class Book(models.Model):
    title = models.CharField(max_length=100)
    authors = models.ManyToManyField(Auther, related_name="books")
    publisher = models.ForeignKey(Publisher)
    publication_date = models.DateField()

    def __str__(self):
        return self.title


class Country(models.Model):
    country = models.CharField(max_length=60)

测试是通过Django日志文件进行的，如何设置请自行google。

1. QuerySet如何工作

Django QuerySet是懒执行的，只有访问到对应数据的时候，才会去访问数据库。
另外如果你再次读取查询到的数据，将不会触发数据库的访问。

例如：请求下面视图，不会进行数据库查询操作，尽管进行了模型类查询

class BookView(View):
    """
    book class
    """
    def get(self, request, *args, **kwargs):
        
        books = models.Book.objects.all()
        
        return HttpResponse("....")

下面对数据进行读取，日志文件中会显示，进行了一次数据库查询。

    def get(self, request, *args, **kwargs):
        books = models.Book.objects.all()
		
        # 此处会读取数据库
        if books:
            print(books)

        book_list = []
        # 再次读取数据，但是不会再访问数据库
        for book in books:
            book_info = {}
            book_info['title'] = book.title
            # book_info['authors'] = book.authors
            book_list.append(book_info)

        return JsonResponse(book_list, safe=False)

(0.001) QUERY = 'SELECT "books_book"."id", "books_book"."title", "books_book"."publisher_id", "books_book"."publication_date" FROM "books_book"' - PARAMS = (); args=()

2. 使用select_related 和 prefetch_related

这两个函数主要是用来优化关联查询的情况，当需要进行关联查询时，如果编写不正确的话，Django 将会多次访问库进行查询，我们应该避免这种情况。尤其是当查询语句位于某个循环中的时候，会导致只需要执行一次的查询重复执行多次。

select_related()

select_related() 通过多表join关联查询，一次性获得所有数据，通过降低数据库查询次数来提升性能，但关联表不能太多，因为join操作本来就比较消耗性能。

publisher是Book模型类的外键，通过外键访问表Publisher获取name，下面有5个Book对象，访问了数据库6次。

def get_book_publisher(request):
    """获取每本书的出版社"""
    books = models.Book.objects.all()

    book_list = []
    for book in books:
        book_info = {}
        book_info[book.title] = book.publisher.name
        book_list.append(book_info)

    return JsonResponse(book_list, safe=False)

(0.001) QUERY = 'SELECT "books_book"."id", "books_book"."title", "books_book"."publisher_id", "books_book"."publication_date" FROM "books_book"' - PARAMS = (); args=()
(0.001) QUERY = 'SELECT "books_publisher"."id", "books_publisher"."name", "books_publisher"."address", "books_publisher"."city", "books_publisher"."state_province", "books_publisher"."country", "books_publisher"."website" FROM "books_publisher" WHERE "books_publisher"."id" = %s' - PARAMS = (3,); args=(3,)
(0.000) QUERY = 'SELECT "books_publisher"."id", "books_publisher"."name", "books_publisher"."address", "books_publisher"."city", "books_publisher"."state_province", "books_publisher"."country", "books_publisher"."website" FROM "books_publisher" WHERE "books_publisher"."id" = %s' - PARAMS = (3,); args=(3,)
(0.000) QUERY = 'SELECT "books_publisher"."id", "books_publisher"."name", "books_publisher"."address", "books_publisher"."city", "books_publisher"."state_province", "books_publisher"."country", "books_publisher"."website" FROM "books_publisher" WHERE "books_publisher"."id" = %s' - PARAMS = (4,); args=(4,)
(0.000) QUERY = 'SELECT "books_publisher"."id", "books_publisher"."name", "books_publisher"."address", "books_publisher"."city", "books_publisher"."state_province", "books_publisher"."country", "books_publisher"."website" FROM "books_publisher" WHERE "books_publisher"."id" = %s' - PARAMS = (4,); args=(4,)
(0.000) QUERY = 'SELECT "books_publisher"."id", "books_publisher"."name", "books_publisher"."address", "books_publisher"."city", "books_publisher"."state_province", "books_publisher"."country", "books_publisher"."website" FROM "books_publisher" WHERE "books_publisher"."id" = %s' - PARAMS = (7,); args=(7,)

使用 select_related 进行优化，对数据库进行了JOIN查询，但是只访问了数据库一次

def get_book_publisher(request):
    """获取每本书的出版社, 优化版"""
    books = models.Book.objects.all().select_related('publisher')

    book_list = []
    for book in books:
        book_info = {}
        book_info[book.title] = book.publisher.name
        book_list.append(book_info)

    return JsonResponse(book_list, safe=False)

(0.001) QUERY = 'SELECT "books_book"."id", "books_book"."title", "books_book"."publisher_id", "books_book"."publication_date", "books_publisher"."id", "books_publisher"."name", "books_publisher"."address", "books_publisher"."city", "books_publisher"."state_province", "books_publisher"."country", "books_publisher"."website" FROM "books_book" INNER JOIN "books_publisher" ON ( "books_book"."publisher_id" = "books_publisher"."id" )' - PARAMS = (); args=()

上面的例子是Book表和Publisher两个表的外键关联，如果此时有第三个表的加入呢，例如新加入一个表Country用来保存出版社的地址，现在我们要查询每本书的出版社地址

def get_book_publisher_country(request):
    """获取每本书的出版社的地址"""
    books = models.Book.objects.all().select_related("publisher")

    book_list = []
    for book in books:
        book_info = (book.title, book.publisher.name, book.publisher.country.country)
        book_list.append(book_info)

    return JsonResponse(book_list, safe=False)

(0.000) QUERY = 'SELECT "books_book"."id", "books_book"."title", "books_book"."publisher_id", "books_book"."publication_date", "books_publisher"."id", "books_publisher"."name", "books_publisher"."address", "books_publisher"."city", "books_publisher"."state_province", "books_publisher"."country_id", "books_publisher"."website" FROM "books_book" INNER JOIN "books_publisher" ON ( "books_book"."publisher_id" = "books_publisher"."id" )' - PARAMS = (); args=()
(0.001) QUERY = 'SELECT "books_country"."id", "books_country"."country" FROM "books_country" WHERE "books_country"."id" = %s' - PARAMS = (3,); args=(3,)
(0.001) QUERY = 'SELECT "books_country"."id", "books_country"."country" FROM "books_country" WHERE "books_country"."id" = %s' - PARAMS = (3,); args=(3,)
(0.000) QUERY = 'SELECT "books_country"."id", "books_country"."country" FROM "books_country" WHERE "books_country"."id" = %s' - PARAMS = (1,); args=(1,)
(0.001) QUERY = 'SELECT "books_country"."id", "books_country"."country" FROM "books_country" WHERE "books_country"."id" = %s' - PARAMS = (1,); args=(1,)
(0.000) QUERY = 'SELECT "books_country"."id", "books_country"."country" FROM "books_country" WHERE "books_country"."id" = %s' - PARAMS = (2,); args=(2,)

因为有了第三个表的加入，获取第三个表的数据时，又访问了5次数据库
继续借助 select_related 进行优化，只不过这次稍微有些不同，涉及到深层查询，需要通过’__’(双下划线)来连接字段，实现递归查询。

下面对原来的内容进行简单修改后，获取数据时，只访问了1次数据库。

def get_book_publisher_country(request):
    """获取每本书的出版社的地址, 优化版"""
    books = models.Book.objects.all().select_related("publisher__country")

    book_list = []
    for book in books:
        book_info = (book.title, book.publisher.name, book.publisher.country.country)
        book_list.append(book_info)

    return JsonResponse(book_list, safe=False)

(0.002) QUERY = 'SELECT "books_book"."id", "books_book"."title", "books_book"."publisher_id", "books_book"."publication_date", "books_publisher"."id", "books_publisher"."name", "books_publisher"."address", "books_publisher"."city", "books_publisher"."state_province", "books_publisher"."country_id", "books_publisher"."website", "books_country"."id", "books_country"."country" FROM "books_book" INNER JOIN "books_publisher" ON ( "books_book"."publisher_id" = "books_publisher"."id" ) INNER JOIN "books_country" ON ( "books_publisher"."country_id" = "books_country"."id" )' - PARAMS = (); args=()

select_related总结

select_related主要针对 ForeignKey 或 OneToOneField 的关系进行优化。
select_related使用SQL的JOIN语句进行优化，通过减少SQL查询的次数来进行优化、提高性能。
当需要深层查询（递归查询）时，可以通过使用双下划线“__”连接字段名来实现指定的递归查询（也就是外键的外键，多层连表查询）。

prefetch_related()

Auther模型类和Book模型类是多对多的关系，通过Book查找每本书的作者，访问了数据库6次

def get_book_authors(request):
    """
    获取每本书的作者
    """
    books = models.Book.objects.all()

    book_list = []
    for book in books:
        book_info = {}
        book_info['book'] = book.title
        book_info['authors'] = [author.first_name for author in book.authors.all()]
        book_list.append(book_info)

    return JsonResponse(book_list, safe=False)

(0.001) QUERY = 'SELECT "books_book"."id", "books_book"."title", "books_book"."publisher_id", "books_book"."publication_date" FROM "books_book"' - PARAMS = (); args=()
(0.000) QUERY = 'SELECT "books_auther"."id", "books_auther"."first_name", "books_auther"."last_name", "books_auther"."email" FROM "books_auther" INNER JOIN "books_book_authors" ON ( "books_auther"."id" = "books_book_authors"."auther_id" ) WHERE "books_book_authors"."book_id" = %s' - PARAMS = (3,); args=(3,)
(0.000) QUERY = 'SELECT "books_auther"."id", "books_auther"."first_name", "books_auther"."last_name", "books_auther"."email" FROM "books_auther" INNER JOIN "books_book_authors" ON ( "books_auther"."id" = "books_book_authors"."auther_id" ) WHERE "books_book_authors"."book_id" = %s' - PARAMS = (4,); args=(4,)
(0.000) QUERY = 'SELECT "books_auther"."id", "books_auther"."first_name", "books_auther"."last_name", "books_auther"."email" FROM "books_auther" INNER JOIN "books_book_authors" ON ( "books_auther"."id" = "books_book_authors"."auther_id" ) WHERE "books_book_authors"."book_id" = %s' - PARAMS = (5,); args=(5,)
(0.001) QUERY = 'SELECT "books_auther"."id", "books_auther"."first_name", "books_auther"."last_name", "books_auther"."email" FROM "books_auther" INNER JOIN "books_book_authors" ON ( "books_auther"."id" = "books_book_authors"."auther_id" ) WHERE "books_book_authors"."book_id" = %s' - PARAMS = (6,); args=(6,)
(0.000) QUERY = 'SELECT "books_auther"."id", "books_auther"."first_name", "books_auther"."last_name", "books_auther"."email" FROM "books_auther" INNER JOIN "books_book_authors" ON ( "books_auther"."id" = "books_book_authors"."auther_id" ) WHERE "books_book_authors"."book_id" = %s' - PARAMS = (7,); args=(7,)

使用prefetch_related()，进行优化，依旧是通过Book查找每本书的作者，但是只访问了数据库两次

def get_book_authors(request):
    """
    获取每本书的作者, 优化版
    """
    books = models.Book.objects.all().prefetch_related('authors')

    book_list = []
    for book in books:
        book_info = {}
        book_info['book'] = book.title
        book_info['authors'] = [author.first_name for author in book.authors.all()]
        book_list.append(book_info)

    return JsonResponse(book_list, safe=False)

(0.004) QUERY = 'SELECT "books_book"."id", "books_book"."title", "books_book"."publisher_id", "books_book"."publication_date" FROM "books_book"' - PARAMS = (); args=()
(0.001) QUERY = 'SELECT ("books_book_authors"."book_id") AS "_prefetch_related_val_book_id", "books_auther"."id", "books_auther"."first_name", "books_auther"."last_name", "books_auther"."email" FROM "books_auther" INNER JOIN "books_book_authors" ON ( "books_auther"."id" = "books_book_authors"."auther_id" ) WHERE "books_book_authors"."book_id" IN (%s, %s, %s, %s, %s)' - PARAMS = (3, 4, 5, 6, 7); args=(3, 4, 5, 6, 7)

prefetch_related 总结

prefetch_related主要针对多对多和多对一关系进行优化。
prefetch_related通过分别获取各个表的内容，然后用Python处理他们之间的关系来进行优化。

小结

主要优化方式还是通过尽可能的减少查询数据库的次数。

pipipippipipipipip

关注

0
点赞
踩
1

收藏

觉得还不错? 一键收藏
0
评论
Django数据库查询优化（一）

数据库查询优化（一）收藏的一篇文章下面内容会使用到的模型类，最近在看TheDjangoBook，这个模型类是里面的，稍微做了些改动# models.pyfrom django.db import models# Create your models here.class Publisher(models.Model): name = models.CharField(...
复制链接

扫一扫