How to Optimize Django ORM

With regards to upgrading an ORM, your most prominent weapon is your comprehension of how your ORM functions in the engine (essentially at a significant level). This makes seeing the principles in general and rules for making an expedient application a lot more straightforward. Accordingly, I enthusiastically suggest perusing the Django docs regarding the matter once. My objective for this post is to consolidate these tips and fools into an effectively referenceable assemblage, just as adding my very own portion. Likewise look at my Django ORM Optimization cheat sheet for a significantly more dense adaptation.

Below are some points to notice on to optimize the ORM

Notice on queryset lazy evalution

Maybe the main part of the Django ORM to comprehend is the manner by which QuerySets work. Since QuerySets are lethargically assessed, you can chain channel() and bar() the entire day without really hitting the data set. Pay special attention to this to assess QuerySets just when you really need to.

When QuerySets are evaluated:

# Iteration
    for person in Person.objects.all():
        # Some logic
    # Slicing/Indexing
    Person.objects.all()[0]
    # Pickling (i.e. serialization)
    pickle.dumps(Person.objects.all())
    # Evaluation functions
    repr(Person.objects.all())
    len(Person.objects.all())
    list(Person.objects.all())
    bool(Person.objects.all())
    # Other
    [person for person in Person.objects.all()]  # List comprehensions
    person in Person.objects.all()  # `in` checks

When QuerySets are not cached:

   # Not reusing evaluated QuerySets
    print([p.name for p in Person.objects.all()])  # QuerySet evaluated and cached
    print([p.name for p in Person.objects.all()])  # New QuerySet is evaluated and cached
    # Slicing/indexing unevaluated QuerySets
    queryset = Person.objects.all()
    print(queryset[0])  # Queries the database
    print(queryset[0])  # Queries the database again
    # Printing
    print(Person.objects.all())

When QuerySets are cached:

   # Reusing an evaluated QuerySet
    queryset = Person.objects.all()
    print([p.name for p in queryset])  # QuerySet evaluated and cached
    print([p.name for p in queryset])  # Cached results are used
    # Slicing/indexing evaluated QuerySets
    queryset = Person.objects.all()
    list(queryset)  # Queryset evaluated and cached
    print(queryset[0])  # Cache used
    print(queryset[0])  # Cache used

Conclusion:

Choosing correct ORM query can minimize the cached memory and leaves the less load on database. It is good practice and also reduce the load time of executing queries.