Django Application consuming memory in the server

Question

I have django application in a Digital Ocean(512MB Memory) with Postgres, Nginx, and Gunicorn on Ubuntu 16.04. On running the application, it consuming more memory. If I navigate through the pages, it also consuming the memory on checking with top command. What will be the problem and what are the possible reason.

Gunicorn

[Unit]
Description=veeyar daemon
After=network.target

[Service]
User=root
Group=www-data
WorkingDirectory=/home/webapps/myproject/
ExecStart=/home/webapps/myproject/venv/bin/gunicorn --access-logfile - --workers 3 --bind unix:/home/webapps/myproject/myproject.sock myproject.wsgi:application

[Install]
WantedBy=multi-user.target

Nginx

server {
    listen 9090;
    location = /favicon.ico { access_log off; log_not_found off; }
    location ^/static/ {
        root /home/webapps/myproject/staticfiles;
    }

    location / {
        include proxy_params;
        proxy_pass http://unix:/home/webapps/myproject/myproject.sock;
    }
}

And also in settings.py I had set DEBUG=False.

I tried by googling it but i cannot understand properly.why it is happening so and did I missed anything. Can you guys please help me to sort out this problem. This will be very great full for me. Thanks in advance.

Answer 1

I will recommend you this post for Django performance, one of the principal reasons of the huge memory in Django is because you are using list instead of iterators.

Regards.

Answer 2

By looking at the memory consumes doesn't really reflect what the problem is or whether there is actually a problem. (unless your concern is how much memory left for other application running from the same box). Otherwise the memory takes up is for caching actually is needed for the application.

I think you should focus on the page loading time (and memory used) instead. For each page are you always retrieving a large queryset and iterate/filter from the code? Are there queryset you can re-use or combined so you don't need to make another query? These are what'd give you trouble when you have more user to the application rather than the memory needed to start the instance.

Answer 3

Check out what the Django docs has to say about application optimisation. It's hard to say what'll be relevant without seeing your code. But it's a good starting point.

One massive help is—if you're iterating over large QuerySets—the use of QuerySet.iterator() - which stops the entire QuerySet from being cached.