減少 django 視圖的執行時間

Question

我有一個 django 視圖，它返回產品 model 的所有產品。 情況是折扣取決於產品和用戶，因此必須在每次運行時計算。 我只添加了 3500 個產品，服務器需要 40-50 秒才能響應。 我想知道添加 100,000 個產品需要多長時間。 我該如何優化呢？

def apply_discount(product_obj,user_obj):
    discount1 = product_obj.brand.discount
    try:
        discount2 = user_obj.special_discount.all().filter(brand=product_obj.brand.name)[0].discount
    except IndexError:
        discount2 = 0
    total_discount = discount1 + discount2
    discount_apply_on = product_obj.brand.discount_on
    price= getattr(product_obj, discount_apply_on)
    final= round(price - (price*total_discount)/100.00,3)
    return (product_obj,price,final,total_discount)


@login_required
def product_list(request):
    context = {}
    product_qs = Product.objects.all()

    product_list = []

    for product in product_qs:
        discount_applied_product = apply_discount(product,request.user)
        product_list.append(discount_applied_product)

    context['products'] = product_list

    return render(request,"myapp/products.html",context)

模型.py

class BrandDiscount(models.Model):
    name = models.CharField(max_length=255)
    discount_on = models.CharField(max_length=25,default="retail_price")
    discount = models.FloatField(default=0)

    def __str__(self):
        return self.name



class Product(models.Model):
    brand = models.ForeignKey(BrandDiscount,on_delete=models.SET_NULL, null=True)
    part_number = models.CharField(max_length=255)
    description = models.TextField(null=True)
    pack_price  = models.FloatField(null=True)
    units_per_pack = models.IntegerField(null=True)
    single_price = models.FloatField(null=True)
    retail_price = models.FloatField(null=True)
    map_price = models.FloatField(null=True)
    jobber_price = models.FloatField(null=True)
    upc = models.CharField(max_length=255)
    stock = models.IntegerField(default=0)
    created_at = models.DateTimeField(auto_now_add=True)

app.yaml

runtime: python37
entrypoint: gunicorn -b :$PORT myProject.wsgi --timeout 120 
instance_class: F4
handlers:
  - url: /static
    static_dir: static/

  - url: /.*
    secure: always
    redirect_http_response_code: 301
    script: auto

我已經在 Google App 引擎標准上部署了網站。 如果我增加 gunicorn 工人，會有幫助嗎？

Answer 1

嘗試對響應進行分頁。 我認為那將是你最好的選擇。

https://docs.djangoproject.com/en/3.0/topics/pagination/

Answer 2

在您的本地環境中，您會在 3 秒內收到包含 400 個元素的響應

如果你做數學

(400 items / 3s) = ~ 133 items/s
(4500 items / 133 items/s) = ~ 33.83s

這種性能與您在 App Engine 中獲得的性能非常相似，正如另一個答案中提到的那樣，您可以對結果進行分頁。

But also you could send a JSON object with the information of all the elements to the interface and let a JavaScript function draw the elements (instead use render to draw all your page) as the user needs them.