如何在Django Debug Toolbar中查看提交/保存时的查询并优化？

Let’s break this down step by step so you can track those queries and optimize your upload_to logic effectively.

1. Ensure Django Debug Toolbar is Set Up Correctly #

First, make sure you’ve got the toolbar properly installed and enabled in your local development environment (it shouldn’t be active in production!). If you haven’t already:

• Install it via pip install django-debug-toolbar
• Add debug_toolbar to your INSTALLED_APPS in settings.py
• Include the toolbar middleware in MIDDLEWARE (place it early in the list, right after django.middleware.common.CommonMiddleware)
• Configure INTERNAL_IPS in settings to allow your local IP access to the toolbar

Once set up, you’ll see the toolbar appear on the right edge of your browser when running your Django app.

2. Track Queries from Submit/Save Operations #

When you submit a form or trigger a model save (including file uploads using upload_to):

1. Complete the submit action as usual (e.g., fill out the form and click "Save").
2. Look for the SQL panel in the Django Debug Toolbar (it’s usually marked with a database icon or labeled "SQL").
3. Click into this panel—you’ll see a full list of all database queries executed during that request, including:
   • Auto-generated ORM queries from Django when saving your model instance
   • Any extra queries you’re running inside your upload_to function

Each query entry will display:

• The full SQL statement
• Execution time
• The number of times it was run (look for text like 1 time or 5 times next to the query—this is exactly what you need to check for repeated calls!)

3. Pinpoint Queries from Your upload_to Logic #

If your upload_to function accesses model attributes that trigger extra database hits (e.g., related objects that aren’t preloaded), those queries will show up in the SQL list. To spot them:

• Look for SELECT queries targeting the related model you’re using in upload_to
• Match the query timing to when the upload_to function runs (it executes right before the file is saved to storage)

For example, if your upload_to looks like this:

def user_directory_path(instance, filename):
    # This might trigger a query if `instance.user` isn't preloaded
    return f'user_{instance.user.id}/{filename}'

You’ll see a SELECT ... FROM auth_user WHERE id = ... query in the SQL panel if the user relation wasn’t fetched with the instance.

4. Confirm Query Call Counts & Optimize #

Once you’ve identified the queries from upload_to:

• Check the execution count: If it says 1 time, your logic is only hitting the database once. If it shows multiple runs, you’re making unnecessary repeated queries.
• Fix repeated queries with these approaches:
  • Preload related objects: When fetching the instance in your view (before saving), use select_related (for foreign keys) or prefetch_related (for many-to-many fields) to load related data upfront. Example:

    # In your view
    instance = MyModel.objects.select_related('user').get(id=request.POST['model_id'])
    instance.save()  # Now upload_to won't trigger an extra query for user
    

  • Cache attribute values: If preloading isn’t feasible, store the needed value in a variable inside upload_to instead of accessing the attribute multiple times. Example:

    def user_directory_path(instance, filename):
        user_id = instance.user.id  # Fetch once, reuse the value
        return f'user_{user_id}/{filename}'
        # Avoid accessing instance.user.id multiple times here
    

内容的提问来源于stack exchange，提问作者user3541631