How To Serve Flask Applications with Gunicorn and Nginx on Ubuntu 14.04

Published on March 21, 2015

Python

Ubuntu

Not using Ubuntu 14.04?

Choose a different version or distribution.

Version

Ubuntu 14.04

Introduction

In this guide, we will be setting up a simple Python application using the Flask micro-framework on Ubuntu 14.04. The bulk of this article will be about how to set up the Gunicorn application server to launch the application and Nginx to act as a front end reverse proxy.

Prerequisites

Before starting on this guide, you should have a non-root user configured on your server. This user needs to have sudo privileges so that it can perform administrative functions. To learn how to set this up, follow our initial server setup guide.

To learn more about the WSGI specification that our application server will use to communicate with our Flask app, you can read the linked section of this guide. Understanding these concepts will make this guide easier to follow.

When you are ready to continue, read on.

Install the Components from the Ubuntu Repositories

Our first step will be to install all of the pieces that we need from the repositories. We will install pip, the Python package manager, in order to install and manage our Python components. We will also get the Python development files needed to build some of the Gunicorn components. We’ll install Nginx now as well.

Update your local package index and then install the packages by typing:

sudo apt-get update
sudo apt-get install python-pip python-dev nginx

Create a Python Virtual Environment

Next, we’ll set up a virtual environment in order to isolate our Flask application from the other Python files on the system.

Start by installing the virtualenv package using pip:

sudo pip install virtualenv

Now, we can make a parent directory for our Flask project. Move into the directory after you create it:

mkdir ~/myproject
cd ~/myproject

We can create a virtual environment to store our Flask project’s Python requirements by typing:

virtualenv myprojectenv

This will install a local copy of Python and pip into a directory called myprojectenv within your project directory.

Before we install applications within the virtual environment, we need to activate it. You can do so by typing:

source myprojectenv/bin/activate

Your prompt will change to indicate that you are now operating within the virtual environment. It will look something like this (myprojectenv)user@host:~/myproject$.

Set Up a Flask Application

Now that you are in your virtual environment, we can install Flask and Gunicorn and get started on designing our application:

Install Flask and Gunicorn

We can use the local instance of pip to install Flask and Gunicorn. Type the following commands to get these two components:

pip install gunicorn flask

Create a Sample App

Now that we have Flask available, we can create a simple application. Flask is a micro-framework. It does not include many of the tools that more full-featured frameworks might, and exists mainly as a module that you can import into your projects to assist you in initializing a web application.

While your application might be more complex, we’ll create our Flask app in a single file, which we will call myproject.py:

nano ~/myproject/myproject.py

Within this file, we’ll place our application code. Basically, we need to import flask and instantiate a Flask object. We can use this to define the functions that should be run when a specific route is requested. We’ll call our Flask application in the code application to replicate the examples you’d find in the WSGI specification:

from flask import Flask
application = Flask(__name__)

@application.route("/")
def hello():
    return "<h1 style='color:blue'>Hello There!</h1>"

if __name__ == "__main__":
    application.run(host='0.0.0.0')

This basically defines what content to present when the root domain is accessed. Save and close the file when you’re finished.

You can test your Flask app by typing:

python myproject.py

Visit your server’s domain name or IP address followed by the port number specified in the terminal output (most likely :5000) in your web browser. You should see something like this:

Flask sample app

When you are finished, hit CTRL-C in your terminal window a few times to stop the Flask development server.

Create the WSGI Entry Point

Next, we’ll create a file that will serve as the entry point for our application. This will tell our Gunicorn server how to interact with the application.

We will call the file wsgi.py:

nano ~/myproject/wsgi.py

The file is incredibly simple, we can simply import the Flask instance from our application and then run it:

from myproject import application

if __name__ == "__main__":
    application.run()

Save and close the file when you are finished.

Testing Gunicorn’s Ability to Serve the Project

Before moving on, we should check that Gunicorn can correctly.

We can do this by simply passing it the name of our entry point. We’ll also specify the interface and port to bind to so that it will be started on a publicly available interface:

cd ~/myproject
gunicorn --bind 0.0.0.0:8000 wsgi

If you visit your server’s domain name or IP address with :8000 appended to the end in your web browser, you should see a page that looks like this:

Flask sample app

When you have confirmed that it’s functioning properly, press CTRL-C in your terminal window.

We’re now done with our virtual environment, so we can deactivate it:

deactivate

Any operations now will be done to the system’s Python environment.

Create an Upstart Script

The next piece we need to take care of is the Upstart script. Creating an Upstart script will allow Ubuntu’s init system to automatically start Gunicorn and serve our Flask application whenever the server boots.

Create a script file ending with .conf within the /etc/init directory to begin:

sudo nano /etc/init/myproject.conf

Inside, we’ll start with a simple description of the script’s purpose. Immediately afterwards, we’ll define the conditions where this script will be started and stopped by the system. The normal system runtime numbers are 2, 3, 4, and 5, so we’ll tell it to start our script when the system reaches one of those runlevels. We’ll tell it to stop on any other runlevel (such as when the server is rebooting, shutting down, or in single-user mode):

description "Gunicorn application server running myproject"

start on runlevel [2345]
stop on runlevel [!2345]

We’ll tell the init system that it should restart the process if it ever fails. Next, we need to define the user and group that Gunicorn should be run as. Our project files are all owned by our own user account, so we will set ourselves as the user to run. The Nginx server runs under the www-data group. We need Nginx to be able to read from and write to the socket file, so we’ll give this group ownership over the process:

description "Gunicorn application server running myproject"

start on runlevel [2345]
stop on runlevel [!2345]

respawn
setuid user
setgid www-data

Next, we need to set up the process so that it can correctly find our files and process them. We’ve installed all of our Python components into a virtual environment, so we need to set an environmental variable with this as our path. We also need to change to our project directory. Afterwards, we can simply call the Gunicorn application with the options we’d like to use.

We will tell it to start 3 worker processes (adjust this as necessary). We will also tell it to create and bind to a Unix socket file within our project directory called myproject.sock. We’ll set a umask value of 007 so that the socket file is created giving access to the owner and group, while restricting other access. Finally, we need to pass in the WSGI entry point file name:

description "Gunicorn application server running myproject"

start on runlevel [2345]
stop on runlevel [!2345]

respawn
setuid user
setgid www-data

env PATH=/home/user/myproject/myprojectenv/bin
chdir /home/user/myproject
exec gunicorn --workers 3 --bind unix:myproject.sock -m 007 wsgi

Save and close the file when you are finished.

You can start the process immediately by typing:

sudo start myproject

Configuring Nginx to Proxy Requests

Our Gunicorn application server should now be up and running, waiting for requests on the socket file in the project directory. We need to configure Nginx to pass web requests to that socket by making some small additions to its configuration file.

Begin by creating a new server block configuration file in Nginx’s sites-available directory. We’ll simply call this myproject to keep in line with the rest of the guide:

sudo nano /etc/nginx/sites-available/myproject

Open up a server block and tell Nginx to listen on the default port 80. We also need to tell it to use this block for requests for our server’s domain name or IP address:

server {
    listen 80;
    server_name server_domain_or_IP;
}

The only other thing that we need to add is a location block that matches every request. Within this block, we’ll include the proxy_params file that specifies some general proxying parameters that need to be set. We’ll then pass the requests to the socket we defined using the proxy_pass directive:

server {
    listen 80;
    server_name server_domain_or_IP;

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

That’s actually all we need to serve our application. Save and close the file when you’re finished.

To enable the Nginx server block configuration we’ve just created, link the file to the sites-enabled directory:

sudo ln -s /etc/nginx/sites-available/myproject /etc/nginx/sites-enabled

With the file in that directory, we can test for syntax errors by typing:

sudo nginx -t

If this returns without indicating any issues, we can restart the Nginx process to read the our new config:

sudo service nginx restart

You should now be able to go to your server’s domain name or IP address in your web browser and see your application:

Flask sample app

Conclusion

In this guide, we’ve created a simple Flask application within a Python virtual environment. We create a WSGI entry point so that any WSGI-capable application server can interface with it, and then configured the Gunicorn app server to provide this function. Afterwards, we created an Upstart script to automatically launch the application server on boot. We created an Nginx server block that passes web client traffic to the application server, relaying external requests.

Flask is a very simple, but extremely flexible framework meant to provide your applications with functionality without being too restrictive about structure and design. You can use the general stack described in this guide to serve the flask applications that you design.

Thanks for learning with the DigitalOcean Community. Check out our offerings for compute, storage, networking, and managed databases.

Learn more about our products

About the author

Justin Ellingwood

Author

See author profile

Former Senior Technical Writer at DigitalOcean, specializing in DevOps topics across multiple Linux distributions, including Ubuntu 18.04, 20.04, 22.04, as well as Debian 10 and 11.

Category:

Tags:

Still looking for an answer?

Ask a question Search for more help

Was this helpful?

This textbox defaults to using Markdown to format your answer.

You can type !ref in this text area to quickly search our full set of tutorials, documentation & marketplace offerings and insert the link!

Ron Smith

April 2, 2015

Don’t you need to activate the virtualenv in the upstart script before running gunicorn?

davecoronel

April 13, 2015

Nice tutorial. I’ve been wanting to understand how WSGI servers worked and also have a chance to try nginx and Flask. I recently learned Bottle and used CGI to serve my app through Apache. So this is like the modern way of doing everything I did earlier. Thanks for that!

lequ

May 7, 2015

When I try to run the “gunicorn --bind 0.0.0.0:8000 wsgi” I get a “Failed to find application : ‘wsgi’”. I’m in the same dir as wsgi.py file. Any ideas?

Thanks in advance!

neekburm

May 15, 2015

I’m having trouble accessing the app after initially configuring it any running the server with python myproject.py. The server is running (It says “Running on http://127.0.0.1:5000/ (Press CTRL+C to quit)”)

When I try to connect using my web browser to XX.XX.XXX.XX:5000, it gives me an unable to connect error. I’m going to continue through the tutorial to setting up gunicorn, and go from there, but what could I be doing wrong?

neekburm

May 15, 2015

Once I ran it with gunicorn, it worked.

sudogaron

June 15, 2015

Following the setup exactly, I get a 502 Bad Gateway error when going to my server IP address. Now I can access the page if I’m running the “gunicorn test:application -b 0.0.0.0:8000” and it loads fine.

I even followed the comment fixes by using the script section instead to activate the env with the right python settings. Still no go.

Has things changed since this was written?

backhousejnr

June 16, 2015

Justin could you please help me with my setup? I have my python app running nginx and gunicorn successfully on windows azure. However, i have a website running on another instance on azure and would like my python app to be served through the website to show something like www.mywebsite.com/nameofmyapp/. I understand its got to do with proxies. Any guide? Thanks.

tomcek112

June 25, 2015

I get an error when I try to run “sudo start myproject”, saying “sudo: start: command not found”. How can I fix this? Thanks for the help!

trimagnus

June 27, 2015

For those who are getting timeouts at the first attempt to connect via port 5000 (or whatever is assigned), I got around the issue by adding that port to the ufw firewall, if you set it up per the initial server setup guide.

This may only help newcomers like myself, but I’m sure there are a few of us around.

matmarsiglio

July 14, 2015

I got and error trying to run sudo start myproject. It says: start: Job failed to start. All the files are exactly like the tutorial. How can I solve this, please?

ktizzel

July 27, 2015

Great writeup thanks a lot! I have a question: How can I set this up so that changes to the Flask application will automatically appear? I tried sudo service nginx restart.

Justin Ellingwood

July 27, 2015

@ktizzel: Try reloading your application server instead:

sudo reload myproject

If that doesn’t work, you can try restarting it:

sudo restart myproject

Hope that helps!

Thanks that worked!

How come it only works if i remove the -m 007 ? edit: It only works perfect if i run the following without (-m 007) directly as the logged in user.

gunicorn --workers 3 --bind unix:myproject.sock  wsgi:app

if i use the above in the upstart script, some pages wont work. If i include -m 007, nothing works. Any ideas?

igorotak6

August 16, 2015

What does include proxy_params do? How can i serve the flask static folder with this setup?

jordantkrueger

September 23, 2015

I’m getting a 502 error. Running gunicorn --bind 0.0.0.0:8000 wsgi:app worked fine.

One deviation from the tutorial is that I cloned in a project I had been working on locally with github, then installed flask and project dependencies via a requirements.txt file, but I don’t think that should cause errors, especially since running gunicorn manually in the virtualenv worked for me.

I tried sudo tail -f /var/log/upstart/flask-portfolio.log (my project is called flask-portfolio) and it spit out this:

    self.manage_workers()
  File "/home/jkrueger/flask-portfolio/portfolio-env/local/lib/python2.7/site-packages/gunicorn/arbiter.py", line 477, in manage_workers
    self.spawn_workers()
  File "/home/jkrueger/flask-portfolio/portfolio-env/local/lib/python2.7/site-packages/gunicorn/arbiter.py", line 541, in spawn_workers
    time.sleep(0.1 * random.random())
  File "/home/jkrueger/flask-portfolio/portfolio-env/local/lib/python2.7/site-packages/gunicorn/arbiter.py", line 214, in handle_chld
    self.reap_workers()
  File "/home/jkrueger/flask-portfolio/portfolio-env/local/lib/python2.7/site-packages/gunicorn/arbiter.py", line 462, in reap_workers
    raise HaltServer(reason, self.APP_LOAD_ERROR)
gunicorn.errors.HaltServer: <HaltServer 'App failed to load.' 4>

I’ve restarted multiple times, and just can’t get this working. I’m very new to this, so I could be missing something obvious.

pguilford

November 8, 2015

After I restart nginx, in the very last step, when I navigate to my domain I’m shown the nginx welcome screen instead of the flask application. All other checkpoints were successful. Any ideas what I could be missing? Thank you for the tutorial.

donohoe

November 24, 2015

This is an AWESOME tutorial - thank you soo much. I follwed it exactly and it worked flawlessly.

joey75d6a5f630e

December 3, 2015

I just wanted to say thank you for this awesome tutorial!

samsonnjogu

December 8, 2015

hello please help me ive been trying to use this tutorial to deploy my application but im getting a strange error (111 connection to upstream client refused) please help ive been stuck for days, when i ru with gunicorn --bind 0.0.0.0:8000 wsgi it works what could be the issue

Ramon Calmet

December 14, 2015

Hello,

I’ve been doing the tutorial as root. When I configure everything I get this error on nginx error.log:

2015/12/14 11:34:59 [crit] 7242#0: *10 connect() to unix:/root/test/test.sock failed (13: Permission denied) while connecting to upstream, client: 62.57.69.106, server: localhost, request: "GET / HTTP/1.1", upst$

I’ve tried to change permissions of test.sock (even 777) and the issue doesn’t disappear. When I start the application with the command gunicorn --bind 0.0.0.0:8000 wsgi inside the folder it works perfectly… Do you have any suggestion?

Thanks

coolit

February 13, 2016

So quick question, and n00b on flask/gunicorn here. What is the purpose of front ending this with nginx if gunicorn can serve up its own webserver?

dariotm

February 26, 2016

Hi!, great tutorial! high quality! Although I am having a problem and I hope you could help. I am using debian (raspbian to be precise). I have gunicorn running with the following command:

  cd ~/workspace/FlaskTestWebapp
  gunicorn --workers 3  --bind unix:testWebApp.sock -m 007 wsgi

And it gives back the pid for the three workers… so it should be fine. I set up the nginx as the guide suggests (I just changed the port):

server {
    listen 8080;
    server_name my-public-domain-name;

    location / {
        include proxy_params;
        proxy_pass http://unix:/home/pi/workspace/FlaskTestWebapp/testWebApp.sock;
    }
}

I have created the link in site-enabled as well. But if I connect to the address I get a “502 Bad Gateway” response…
I have tried several things already, please help!

vasaulys

February 29, 2016

The -m flag was throwing me off! I used -m 500 and it worked fine.

This flag in gunicorn is used for umask, which is a linux utility that I only half-understand. You can read more about it here (which, while written for Arch, is applicable for both Arch & Debian-based systems) https://wiki.archlinux.org/index.php/Umask

tkiran05

September 16, 2016

sudo start myproject throws start: Unknown job: myproject

redmondinho

November 3, 2016

Great tutorial. I have a single Flask app running using the above. I am struggling to get multiple apps running though or getting the flask app served from a location other than /. I’m trying something like this;

server { listen 80; server_name _;

location /app1 {
    include proxy_params;
    proxy_pass http://unix:/home/user/myproject1/myproject1.sock;
}

location /app2 {
    include proxy_params;
    proxy_pass http://unix:/home/user/myproject1/myproject2.sock;
}

}

I have the 2 apps set up and they both work individually when the location in sites-available is / but can’t get the app running from another location, which was my first step to getting multiple apps running.

Any help would be appreciated.

elmq0022

December 20, 2016

Thanks for the tutorial. I have one question though. What port do I open with ufw so pip will continue to work as normal?

I saw some stuff on the web saying port 8123/tcp, but I didn’t have any luck with that. Any help is greatly appreciated.

angshumanghosh1234

January 5, 2017

I got one error over and over again while trying to set up nginx and flask on a virtual machine. I copy pasted everything, but I still just saw the default nginx webpage. To fix it, I finally realized that I had to go to /etc/nginx/sites-available and delete the default config, and then it worked perfectly

cpante

February 1, 2017

hello im using debian, sudo start, start command not found?

aitomon

February 4, 2017

Hello everyone, I am using ubuntu 16.10 machine, and when I try to execute sudo start myproject, it tells me start command not found. I googled for some time and found out that ubuntu 16.10 uses systemd and not upstart, so how am I suppose to execute above command for 16.10 machine ?

anioko1

April 24, 2017

@jellingwood I have this error “init: Failed to spawn myproject main process: unable to find setuid user”. Here is myproject.conf :

description “Gunicorn application server running myproject”

start on runlevel [2345] stop on runlevel [!2345]

respawn setuid user setgid www-data

env PATH=/root/myproject/myprojectenv/bin chdir /root/myproject exec gunicorn --workers 3 --bind unix:myproject.sock -m 007 wsgi

I dont have home/user/myproject/ instead I have root/myproject and I also changed myproject.py into app.py . I had it running on port 5000 and port 8000.

These are my error logs:

random: nonblocking pool is initialized Disconnected from Upstart Disconnected from Upstart init: upstart-udev-bridge main process (667) terminated with status 1 init: upstart-udev-bridge main process ended, respawning init: upstart-socket-bridge main process (953) terminated with status 1 init: upstart-socket-bridge main process ended, respawning init: upstart-file-bridge main process (1261) terminated with status 1 init: upstart-file-bridge main process ended, respawning (root) CMD ( cd / && run-parts --report /etc/cron.hourly) init: Failed to spawn myproject main process: unable to change working directory: No such file or directory init: Failed to spawn myproject main process: unable to change working directory: No such file or directory init: Failed to spawn myproject main process: unable to change working directory: No such file or directory Failed to spawn myproject main process: unable to change working directory: No such file or directory init: Failed to spawn myproject main process: unable to find setuid user init: Failed to spawn myproject main process: unable to find setuid user

Chris Wilkins

May 23, 2017

Down at the step of creating the .conf file in the /etc/init directory it seems there are some important parts missing in the conf file. I suspect the pathing.

When I run ‘sudo start (myproject)’ I get:

“sudo: start: command not found”

Is the syntax in:

exec gunicorn --workers 3 --bind unix:myproject.sock -m 007 wsgi

correct? Should there be more comprehensive paths in here?

Thanks,

Chris.

Ananda Theerthan

September 3, 2017

Thanks for this great tutorial, while try to access the domain i get below error. 017/09/03 06:25:18 [error] 15994#0: *1 connect() to unix:/home/ubuntu/4k/smilesnaturecure/mysnc.sock failed (111: Connection refused) while connecting to upstream, client: 182.65.222.167, server: smilesnaturecure.com, request: “GET /favicon.ico HTTP/1.1”, upstream: “http://unix:/home/ubuntu/4k/smilesnaturecure/mysnc.sock:/favicon.ico”, host: “www.smilesnaturecure.com”, referrer: “http://www.smilesnaturecure.com/”

mihinsumaria

January 26, 2019

Hi! The upstart script needs sudo privileges as you mentioned. I do not have sudo access, and I installed all of the dependencies from source. How is it possible to have an upstart script without root privileges?