nginx error "conflicting server name" ignored

Question

server {
    #listen   80; ## listen for ipv4; this line is default and implied
    #listen   [::]:80 default ipv6only=on; ## listen for ipv6

    #root /usr/share/nginx/www;

root /home/ubuntu/node-login;
    # Make site accessible from 
    server_name ec2-xx-xx-xxx-xxx.us-west-1.compute.amazonaws.com;

location /{
    proxy_pass http://127.0.0.1:8000/;
    proxy_redirect off;
}
}

this results in nignx error [warn] conflicting server name "ec2..." on 0.0.0.0:80 ignored I dont understand, any explanation appreciated. Thanks.

There seems to be an implicit default 0.0.0.0:80, adding it explicitly worked for me. listen ec2-xx-xx-xxx-xxx.us-west-1.compute.amazonaws.com:80; followed by the server_name directive worked. — user1447121, Jul 11 '12 at 17:24
Check `/etc/nginx/sites-available/` that is linked to `/etc/nginx/sites-enabled/`. Additional files here may contain duplicate or conflicting configuration. — Hanxue, Dec 23 '15 at 10:36
changing `config.force_ssl = true` to `false` solved my problem, this took me days to figure out — Ibukun Muyide, Jul 06 '18 at 02:02
For me this same error was from a missing semi-colon at the end of one of the `server_name` lines when I had several `server` entries. — Kevin, May 24 '19 at 12:04
Certbot from LetsEncrypt had added a server block in the default file in my case. It did that cause I tried to get the certificate before I had setup the separate server block. — Alex D, Oct 25 '19 at 17:46
In my case I had [multiple](https://stackoverflow.com/a/51480365/2195238) `server` block in one file. After I separated them to ist own config files warnings were gone. — Jasom Dotnet, Dec 27 '20 at 08:47
If you are missing the `listen` directive it can cause the same error to show. — Obay Abd-Algader, Jun 01 '22 at 08:36
I dumbly got this error when omitting the semi-colon at the end of the server_name line. — Paul, Apr 18 '23 at 17:53

score 215 · Accepted Answer · edited May 23 '17 at 12:26

215

I assume that you're running a Linux, and you're using gEdit to edit your files. In the /etc/nginx/sites-enabled, it may have left a temp file e.g. default~ (watch the ~).

Depending on your editor, the file could be named .save or something like it. Just run $ ls -lah to see which files are unintended to be there and remove them (Thanks @Tisch for this).

Delete this file, and it will solve your problem.

edited May 23 '17 at 12:26

Community

1
1

answered Dec 23 '12 at 13:45

Omar Al-Ithawi

4,988
5
36
47

fixed the issue for me on Ubuntu 12.04 x86_64 + nginx – Stephen Sprinkle Jan 10 '13 at 05:16
2

can you clarify? I don't see any "default~" files at all in my /etc/nginx/sites-enabled only the "default" folder. – David Saintloth Jan 10 '14 at 05:27
@DavidSaintloth it worked for me and the OP. This is one source of problem, it could be anything else. But to make sure you really don't have the file do: `$ ls -alF /etc/nginx/sites-enabled` normally it's not shown unless you specifically ask for it since it's a hidden/tmp file. – Omar Al-Ithawi Jan 11 '14 at 17:09
@DavidSaintloth Check the logs if that don't work, ask a separate question. – Omar Al-Ithawi Jan 11 '14 at 17:10
.... where is this `default~` file located? – Jake Wilson Jun 29 '15 at 02:35
3

@Jakobud in `/etc/nginx/sites-enabled`. I have edited my answer to mention it. – Omar Al-Ithawi Jul 02 '15 at 10:07
In my case nginx.conf had the include for the sites-enabled folder twice. removing one solved the problem – Fabian de Pabian Jul 21 '15 at 20:03
5

I had a .save file that was causing the issue. – Tisch Feb 01 '16 at 11:37
n had removed backup files (`default~`) from `sites-available` too – jerinisready Jan 19 '17 at 19:40
2

In my case some configuration of emacs created a file with a leading and trailing `#`, e.g. `#portal#`. – taffit Feb 02 '17 at 08:04
Thanks, this solved my problem! I had followed a tutorial and did not realize keeping the "hello world" file in /etc/nginx/sites-available would be an issue. I removed it and default, only keeping the file I wanted to deploy and that fixed my issue. – ben.dichter Apr 04 '17 at 18:00
1

Nice you can do `ls -la /etc/nginx/sites-enabled/`end then `rm /etc/nginx/sites-enabled/default` **be careful dont delete** – Marco Feregrino Jul 11 '18 at 01:22

score 41 · Answer 2 · answered Jul 11 '12 at 12:24

41

You have another server_name ec2-xx-xx-xxx-xxx.us-west-1.compute.amazonaws.com somewhere in the config.

answered Jul 11 '12 at 12:24

VBart

14,714
4
45
49

2

I checked for that and didn't find any. it does not give this warning when i remove the "ec2_xxx ...". Is there an implicit default for 0.0.0.0:80 in nginx config ? – user1447121 Jul 11 '12 at 17:06
1

[Official documentation](http://nginx.org/r/listen): `default: listen *:80 | *:8000;` – VBart Oct 08 '12 at 20:06
2

This was it for me, however it was `localhost`. – tim-phillips Nov 20 '14 at 19:22
@VBart your link gave me a hint, actually my browser was prefixing https instead of http in my url, thanks – Nikhil Bhardwaj Jan 06 '22 at 08:23
1

I used: cd /etc/nginx/sites-available; grep -rnw . -e domain.com – Alex Jan 08 '23 at 21:44
In my case certbot added duplicate configuration into default configuration file. just remove the configuration from default config file located at /etc/nginx/conf.d/default.conf: – Roman M Jul 30 '23 at 19:23

score 23 · Answer 3 · answered Mar 06 '13 at 01:32

23

There should be only one localhost defined, check sites-enabled or nginx.conf.

answered Mar 06 '13 at 01:32

Marin

12,531
17
56
80

12

I had two identical "server_name" directives in two separate *.conf files. Thanks. – Felipe Alvarez Apr 05 '16 at 04:24
3

This solved my issue as well. I had placed 2 same server blocks in different *.conf files. Thanks! – Pranav Jituri May 11 '17 at 10:25
2

Also, check all files in this directory: `/etc/nginx/conf.d`. Each domain can appear only once (as a `servername` value) amongst files in that directory, as well. That was my issue! – SherylHohman Nov 13 '18 at 22:24

nginx error "conflicting server name" ignored

3 Answers3

Linked