Django/Apache/mod_wsgi Deployment Issue: Internal Server Error & Redirect Loop on cPanel (AlmaLinux 9)

Question

I'm an intern trying to deploy a Django application on a cPanel server, and I've run into a persistent "Internal Server Error" coupled with an Apache redirect loop. I've been troubleshooting for over two days with various AI helpers, but I can't pinpoint the exact cause. Any fresh perspectives ??

---

My Setup:

• Server OS : AlmaLinux 9.6 (Sage Margay)
• Hosting Environment: cPanel (My cPanel account does not have the "Setup Python App" option, which is why I'm using a manual Apache/mod_wsgi configuration.)
• Django Deployment: Apache + mod_wsgi
• No .htaccess content at all.
• Project Structure: I have two Django instances (production and staging) for CI/CD,

each with its own settings, environment variables, database, and domain.

• validator/settings/base.py

• validator/settings/production.py (inherits from base)

• validator/settings/staging.py (inherits from base)

• Environment files are stored in /etc/validator/envs

• WSGI files are stored in /etc/validator/wsgi/

• Django project roots: /home/<username>/<app_name>/ (prod) and /home/<username>/<app_name_test>/ (staging).

• Python Version: My virtual environments are running Python 3.12.9, and my WSGI files assert this.

---

The Problem:

When I try to access either domain.in (production) or test.domain.in (staging), I receive an "Internal Server Error". My Apache error log (specifically /var/log/httpd/error_log) shows the following critical error message:

[Fri May 30 00:26:38.624027 2025] [core:error] [pid 253617:tid 253705] [client 45.115.89.80:26970] AH00124: Request exceeded the limit of 10 internal redirects due to probable configuration error. Use 'LimitInternalRecursion' to increase the limit if necessary. Use 'LogLevel debug' to get a backtrace.

This started after a server expiration/renewal and then when I configured new databases for staging and production, updating the respective .env files. Previously, it worked (sharing a database).

---

My Apache configuration:

I understand cPanel uses a "Direct Include Style" for Apache configs, where httpd.conf includes files from directories like /etc/apache2/conf.d/userdata/.... My custom Django configurations are placed in these included directories.

1. Production Domain Custom Apache Config (/etc/apache2/conf.d/userdata/std/2_4/<username>/<domain.in>/<filename>.conf):

ServerName <domain>.in
ServerAlias www.<domain>.in
ServerAlias <server_ip>
WSGIDaemonProcess django_app python-home=/home/<username>/app/venv python-path=/home/<username>/app user=<apache_user> group=<group_user>
WSGIProcessGroup app
WSGIScriptAlias / /etc/validator/wsgi/production.wsgi 
Alias /static/ /home/<username>/app/staticfiles/
<Directory /home/<username>/app/staticfiles>
 Require all granted
</Directory>
<Directory /etc/validator/wsgi>
 <Files production.wsgi>
 Require all granted
 </Files>
</Directory>
ErrorLog /home/<username>/app/error.log
CustomLog /home/<username>/app/access.log combined

2. Staging Domain Custom Apache Config (similar path for test.domain.in):

ServerName test.<domain>.in
ServerAlias www.test.<domain>.in
WSGIDaemonProcess test_app python-home=/home/<username>/app_test/venv python-path=/home/<username>/app_test user=<apache_user> group=<group_user>
WSGIProcessGroup app_test
WSGIScriptAlias / /etc/validator/wsgi/staging.wsgi 
Alias /static/ /home/<username>/app_test/staticfiles/
<Directory /home/<username>/app_test/staticfiles>
 Require all granted
</Directory>
<Directory /etc/validator/wsgi>
 <Files production.wsgi>
 Require all granted
 </Files>
</Directory>
ErrorLog /home/<username>/app_test/error.log
CustomLog /home/<username>/app_test/access.log combined

3.Production Environment File (/etc/validator/envs/production.env):

DBUSER=<dbuser>
DBPASSWORD=<dbpasswd>
DBNAME=<dbname>
DBHOST=localhost
DBPORT=3306

4. Production WSGI File (/etc/validator/wsgi/production.wsgi):

import os
import sys
from pathlib import Path
from dotenv import load_dotenv
ENV_FILE = Path('/etc/validator/envs/production.env')
load_dotenv(dotenv_path=ENV_FILE)
os.environ.setdefault('DJANGO_SETTINGS_MODULE', '<django_root_app>.settings.production')
from django.core.wsgi import get_wsgi_application
application = get_wsgi_application()

5. Django settings/prod.py (similar structure for settings/staging.py):

import os
from .base import *
DEBUG = False
ALLOWED_HOSTS = [
 "domain",
 "www.domain",
 "server_ip",
]
DATABASES = {
 "default": {
 "ENGINE": "django.db.backends.<sql>",
 "USER": os.environ["DBUSER"],
 "PASSWORD": os.environ["DBPASSWORD"],
 "NAME": os.environ["DBNAME"],
 "HOST": os.environ["DBHOST"],
 "PORT": os.environ["DBPORT"],
 }
}

---

Troubleshooting Steps I've Already Taken:

1. Checked Apache Error Logs: The main Apache log (/var/log/httpd/error_log) consistently shows the "Request exceeded the limit of 10 internal redirects" error. My Django app's custom logs (/home//app/error.log and _test/error.log) are empty or show no relevant errors, often indicating a problem before Django's logger can even kick in.

2. Permissions: I've meticulously checked and set chown -R medivali:webapps and chmod permissions (755 for directories, 644 for files, 664 for logs) on /home//app/, /home//app_test/, /etc/validator/envs/, and /etc/validator/wsgi/.

3. WSGIDaemonProcess User/Group: I've tried user=nobody group=webapps (as in config) and user=medivali group=medivali. Neither resolves the issue.

4. Python Version Assertion: Confirmed venv/bin/python --version is 3.12+ for both environments.

5. proxy_fcgi_module Conflict: I've added <FilesMatch ".(phtml|php[0-9]*)$">SetHandler None to my validator_conf.conf files to try and prevent the cPanel PHP-FPM handler from interfering.

6. DocumentRoot vs. WSGIScriptAlias: I understand WSGIScriptAlias / should override DocumentRoot for Django, but I've also tried creating an index.html in /home/medivali/public_html (the default DocumentRoot in the main VHost) to test basic Apache serving, but it still leads to the redirect loop or internal server error.

7. Database Credentials: I've triple-checked the DBUSER, DBPASSWORD, DBNAME, DBHOST, DBPORT in both .env files. I can connect to the databases directly from the server using mysql client.

8. ALLOWED_HOSTS: Confirmed correct domain names are in ALLOWED_HOSTS for both production and staging settings.

9. AI Consultations: Spent 2.5 days consulting ChatGPT, Gemini, Claude, and DeepSeek, trying various suggestions, but without success.

• Airith: Try setting LogLevel debug to see if there's more logs. This is generally caused by a RewriteRule somewhere in your apache config.
• Adnan: Thanks for the suggestion! ’ve checked my .htaccess and main Apache configs, and I don’t see any RewriteRule directives defined anywhere. I’ll keep digging :)

Answer 1

It was a Python build issue. I figured it out — some C-extension modules (like for pickle) were not properly built, which caused the internal server error. After rebuilding the environment properly, the issue was resolved. Thank you!