Nginx config can't reload. Have "bind() to 0.0.0.0:80 failed (98: Address already in use)" errors [duplicate]

Question

I tried installing an HTTPS certificate, and after banging my head against the wall for a few hours, I realized that Nginx is NOT picking up the changes I made in my site configuration files contained within /etc/nginx/sites-available/

When I run service nginx restart it restarts fine. When I run nginx -t it says my syntax is OK and my config test is successful. However, when I run service nginx reload I get an error that says reload: Not running

When I look in /var/log/nginx/error.log I have these errors.

2016/08/29 16:04:40 [emerg] 14744#14744: bind() to 0.0.0.0:80 failed (98: Address already in use)
2016/08/29 16:04:40 [emerg] 14744#14744: bind() to 0.0.0.0:443 failed (98: Address already in use)
2016/08/29 16:04:40 [emerg] 14744#14744: bind() to [::]:80 failed (98: Address already in use)
2016/08/29 16:04:40 [emerg] 14744#14744: bind() to 0.0.0.0:80 failed (98: Address already in use)
2016/08/29 16:04:40 [emerg] 14744#14744: bind() to 0.0.0.0:443 failed (98: Address already in use)
2016/08/29 16:04:40 [emerg] 14744#14744: bind() to [::]:80 failed (98: Address already in use)
2016/08/29 16:04:40 [emerg] 14744#14744: bind() to 0.0.0.0:80 failed (98: Address already in use)
2016/08/29 16:04:40 [emerg] 14744#14744: bind() to 0.0.0.0:443 failed (98: Address already in use)
2016/08/29 16:04:40 [emerg] 14744#14744: bind() to [::]:80 failed (98: Address already in use)
2016/08/29 16:04:40 [emerg] 14744#14744: bind() to 0.0.0.0:80 failed (98: Address already in use)
2016/08/29 16:04:40 [emerg] 14744#14744: bind() to 0.0.0.0:443 failed (98: Address already in use)
2016/08/29 16:04:40 [emerg] 14744#14744: bind() to [::]:80 failed (98: Address already in use)
2016/08/29 16:04:40 [emerg] 14744#14744: bind() to 0.0.0.0:80 failed (98: Address already in use)
2016/08/29 16:04:40 [emerg] 14744#14744: bind() to 0.0.0.0:443 failed (98: Address already in use)
2016/08/29 16:04:40 [emerg] 14744#14744: bind() to [::]:80 failed (98: Address already in use)
2016/08/29 16:04:40 [emerg] 14744#14744: still could not bind()

After Googling around, I ran across this command netstat -plutn | grep 80 which gives me...

tcp        0      0 0.0.0.0:80              0.0.0.0:*               LISTEN      14642/nginx: worker
tcp6       0      0 :::80                   :::*                    LISTEN      14642/nginx: worker

I have no idea what this really means, but reading around, it seems Nginx should NOT be running on port 80? Nginx is my webserver, don't I want it running on port 80?

Not sure how to proceed. I want Nginx to pick up the changes I made in my site config files with the service nginx reload command, and at the same time I'm hesitant to just copy/paste commands I find around the net since I have 10 websites hosted on this server.

Where am I going wrong? I have tried deleting the default file in my sites-available folder as some have suggested, but I just don't really understand what the problem is, and how to fix it.

Answer 1

Problem is fixed. As Tim said, I ran the ps -ef | grep nginx command and manually ran kill PID for all the processes. Then I ran service nginx restart and all worked well. When I then ran service nginx reload I did NOT get an error, and my HTTPS cert now works. Thanks Tim and dmourati for your assistance

Answer 2

Sounds like a problem with your init script and the pid file location for nginx. You can manually kill nginx like so:

sudo kill 14642

Then start nginx via the init script as follows:

sudo service nginx start

From there, if you copy/paste your init script and nginx.conf file, I'm sure we can find the problem with the pid.