Deployment

Isso ships with a built-in web server, which is useful for the initial setup and may be used in production for low-traffic sites (up to 20 requests per second). Running a “real” WSGI server supports nice things such as UNIX domain sockets, daemonization and solid HTTP handler. WSGI servers are more stable, secure and web-scale than the built-in web server.

gevent

Probably the easiest deployment method. Install with PIP (requires libevent):

$ pip install gevent

Then, just use the isso executable as usual. Gevent monkey-patches Python’s standard library to work with greenlets.

To execute Isso, just use the commandline interface:

$ isso -c my.cfg run

Unfortunately, gevent 0.13.2 does not support UNIX domain sockets (see #295 and #299 for details).

uWSGI

Isso has special support for uWSGI, namely fast IPC caching, job spooling and delayed jobs. It is the author’s choice, but not the only one. You need uWSGI 1.9 or higher, fortunately you can install it from PyPi:

~> apt-get install build-essential python-dev
~> pip install uwsgi

For convenience, I recommend a INI-style configuration (you can also supply everything as command-line arguments):

[uwsgi]
http = :8080
master = true
; set to `nproc`
processes = 4
cache2 = name=hash,items=1024,blocksize=32
; you may change this
spooler = /tmp/isso/mail
module = isso.run
; uncomment if you use a virtual environment
; virtualenv = /path/to/isso
env = ISSO_SETTINGS=/path/to/isso.cfg

Then, create the spooling directory and start Isso via uWSGI:

~> mkdir /tmp/isso/mail
~> uwsgi /path/to/uwsgi.ini

gunicorn

Gunicorn ‘Green Unicorn’ is a Python WSGI HTTP Server for UNIX with a pre-fork worker ported from Ruby’s Unicorn project. Install gunicorn via PIP:

$ pip install gunicorn

To execute Isso, use a command similar to:

$ export ISSO_SETTINGS="/path/to/isso.cfg"
$ gunicorn -b localhost:8080 -w 4 --preload isso.run

mod_wsgi

First, create a startup script, called isso.wsgi. If Isso is in your system module search path, then the script is quite simple. This script is included in the isso distribution as run.py:

import os

from isso import make_app
from isso import dist, config

application = make_app(
config.load(
    config.default_file(),
    "/path/to/isso.cfg"),
multiprocessing=True)

If you have installed Isso in a virtual environment, then you will have to add the path of the virtualenv to the site-specific paths of Python:

  import site
  site.addsitedir("/path/to/isso_virtualenv")

  import os

  from isso import make_app
  from isso import dist, config

  application = make_app(
  config.load(
      config.default_file(),
      "/path/to/isso.cfg",
  multiprocessing=True)

Using the aforementioned script will load system modules when available and modules from the virtualenv otherwise. Should you want the opposite behavior, where modules from the virtualenv have priority over system modules, the following script does the trick:

  import os
  import site
  import sys

  # Remember original sys.path.
  prev_sys_path = list(sys.path)

  # Add the new site-packages directory.
  site.addsitedir("/path/to/isso_virtualenv")

  # Reorder sys.path so new directories at the front.
  new_sys_path = []
  for item in list(sys.path):
      if item not in prev_sys_path:
          new_sys_path.append(item)
          sys.path.remove(item)
  sys.path[:0] = new_sys_path

  from isso import make_app
  from isso import dist, config

  application = make_app(
  config.load(
      config.default_file(),
      "/path/to/isso.cfg",
  multiprocessing=True)

The last two scripts are based on those given by mod_wsgi documentation.

The Apache configuration will then be similar to the following:

<VirtualHost *>
    ServerName example.org

    WSGIDaemonProcess isso user=www-data group=www-data threads=5
    WSGIScriptAlias /mounted_isso_path /path/to/isso.wsgi
</VirtualHost>

You will need to adjust the user and group according to your Apache installation and security policy. Be aware that the directory containing the comments database must be writable by the user or group running the WSGI daemon process: having a writable database only is not enough, since SQLite will need to create a lock file in the same directory.

mod_fastcgi

You can use this method if your hosting provider doesn’t allow you to have long running processes. If FastCGI has not yet been configured in your server, please follow these steps:

Note

This information may be incorrect, if you have more knowledge on how to deploy Python via mod_fastcgi, consider extending/correcting this section.

For more information, see Flask: Configuring Apache.

LoadModule fastcgi_module /usr/lib64/httpd/modules/mod_fastcgi.so

FastCgiServer /var/www/html/yourapplication/app.fcgi -idle-timeout 300 -processes 5

<VirtualHost *>
    ServerName example.org

    AddHandler fastcgi-script fcgi
    ScriptAlias / /var/www/isso.fcgi

    <Location />
        SetHandler fastcgi-script
    </Location>
</VirtualHost>

Next, to run isso as a FastCGI script you’ll need to install flup with PIP:

$ pip install flup

Finally, copy’n’paste to /var/www/isso.fcgi (or whatever location you prefer):

#!/usr/bin/env python
#: uncomment if you're using a virtualenv
# import sys
# sys.path.insert(0, '<your_local_path>/lib/python3.<ver>/site-packages')

from isso import make_app, dist, config
import os

from flup.server.fcgi import WSGIServer

application = make_app(config.load(
    config.default_file(),
    "/path/to/isso.cfg"))
WSGIServer(application).run()