Usinv virtualenv in django image?

Question

Hello!

I’ve used Django on Ubuntu image to create my droplet and it all worked well until I’ve tried to use virtualenv for django_project dependencies.

I’ve created home/django/venvs dir. Within it I’ve created venv that I have installed my app’s dependencies to.

Then I’ve edited upstart file to following:

description "Gunicorn daemon for Django project"

start on (local-filesystems and net-device-up IFACE=eth0)
stop on runlevel [!12345]

# If the process quits unexpectadly trigger a respawn
respawn

setuid django
setgid django
chdir /home/django

script
    source venvs/myapp/bin/activate
end script

exec gunicorn \
    --name=django_project \
    --pythonpath=django_project \
    --bind=0.0.0.0:9000 \
    --config /etc/gunicorn.d/gunicorn.py \
    django_project.wsgi:application

However Gunicorn fails to start app, which crashes on import of module thats in venv, but not in global python.

Question is what I am doing wrong? How to make default django_project use venv?

jeffree · Answer

I am having a similar issue I have a django project(forte) running from a virtualenv(prod)

If I

root@Fortelinea:/# cd /home/django/forte/
root@Fortelinea:/home/django/forte# /home/django/prod/bin/gunicorn forte.wsgi -b 127.0.0.1:9000
[2015-02-04 18:47:02 -0500] [1339] [INFO] Starting gunicorn 19.2.0
[2015-02-04 18:47:02 -0500] [1339] [INFO] Listening at: http://127.0.0.1:9000 (1339)
[2015-02-04 18:47:02 -0500] [1339] [INFO] Using worker: sync
[2015-02-04 18:47:02 -0500] [1342] [INFO] Booting worker with pid: 1342

then everything is groovy and works

now I am trying to use upstart to lauch my virtualenv gunicorn at boot

my ustart script looks like (/etc/init/forte.conf)


start on (local-filesystems and net-device-up IFACE=eth0)
stop on runlevel [!12345]

# If the process quits unexpectadly trigger a respawn
respawn

setuid django
setgid django
chdir /home/django/forte

exec /home/django/prod/bin/gunicorn forte.wsgi -b 127.0.0.1:9000

when this runs it the process fails and spawns a bunch of time and then quits

root@Fortelinea:/# start forte
forte start/running, process 1359
root@Fortelinea:/# tail /var/log/syslog
Feb  4 18:52:37 Fortelinea kernel: [ 1760.050336] init: forte main process (1389) terminated with status 3
Feb  4 18:52:37 Fortelinea kernel: [ 1760.050350] init: forte main process ended, respawning
Feb  4 18:52:38 Fortelinea kernel: [ 1760.431913] init: forte main process (1394) terminated with status 3
Feb  4 18:52:38 Fortelinea kernel: [ 1760.431927] init: forte main process ended, respawning
Feb  4 18:52:38 Fortelinea kernel: [ 1760.815041] init: forte main process (1399) terminated with status 3
Feb  4 18:52:38 Fortelinea kernel: [ 1760.815057] init: forte main process ended, respawning
Feb  4 18:52:39 Fortelinea kernel: [ 1761.194726] init: forte main process (1404) terminated with status 3
Feb  4 18:52:39 Fortelinea kernel: [ 1761.194741] init: forte main process ended, respawning
Feb  4 18:52:39 Fortelinea kernel: [ 1761.577727] init: forte main process (1409) terminated with status 3
Feb  4 18:52:39 Fortelinea kernel: [ 1761.577744] init: forte respawning too fast, stopped

any thoughts or insights would be appreciated

rafalpiton · Answer

Thank you for your attention Andrew.

I think I’ve got it fixed, but there’s one gotcha:

For some reason latest gunicorn that worked for me was 17.5 (the one that came preinstaled with image).

I’ve got app to work, but there are still quirks, which I will solve myself now (fingers crossed)

Andrew SB · Answer

The script block executes on its own. So the virtualenv will not be available to gunicorn when started like that. Check gunicorn’s docs on using it with a virtualenv: To serve an app from a Virtualenv it is generally easiest to just install Gunicorn directly into the Virtualenv. This will create a set of Gunicorn scripts for that Virtualenv which can be used to run applications normally. So install it directly in the virtualenv: source ~/venvs/myapp/bin/activate pip install -I gunicorn Then in your Upstart script, you’d need to use gunicorn from the virtualenv. It would be something like: description "Gunicorn daemon for Django project" start on (local-filesystems and net-device-up IFACE=eth0) stop on runlevel [!12345] # If the process quits unexpectadly trigger a respawn respawn setuid django setgid django chdir /home/django exec venvs/myapp/bin/gunicorn \ --name=django_project \ --pythonpath=django_project \ --bind=0.0.0.0:9000 \ --config /etc/gunicorn.d/gunicorn.py \ django_project.wsgi:application