Otdfbt

I have asked a related question but has solved it already.

Using nginx provided a reverse proxy.

The config file is taking from project document, I paste it below.

Current situation is:

The backend service is working on port 8080. And the front end is working on port 9001.

$ netstat -anp |grep 8080
(Not all processes could be identified, non-owned process info
 will not be shown, you would have to be root to see it all.)
tcp6 0 0 :::8080 :::* LISTEN 26707/java

$ netstat -anp |grep 9001
(Not all processes could be identified, non-owned process info
 will not be shown, you would have to be root to see it all.)
tcp 0 0 127.0.0.1:9001 0.0.0.0:* LISTEN 12130/grunt

I can visit the localhost:9001 on machine A where the service is running.
While visits from any other machine in the same LAN. It shows the error page. If go with Inspect, there is a message in the error page.

Cannot open document for: /eplmp-server-rest/api/auth/providers from http://192.168.1.164:8989/

192.168.1.164 is the machine A's IP address.
/eplmp-server-rest/api/* should be the endpoint serve by the backend service.

The XHR Network Headers from machine A.

General
Request URL: http://localhost:8080/eplmp-server-rest/api/auth/providers
Request Method: GET
Status Code: 200 OK
Remote Address: [::1]:8080
Referrer Policy: no-referrer-when-downgrade

Requst Headers
Provisional headers are shown
Accept: application/json, text/javascript, */*; q=0.01
Origin: http://localhost:9001
Referer: http://localhost:9001/
User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/73.0.3683.103 Safari/537.36

from another machine in the same LAN

Request URL: http://localhost:8080/eplmp-server-rest/api/auth/providers
Referrer Policy: no-referrer-when-downgrade

Provisional headers are shown
Accept: application/json, text/javascript, /; q=0.01
Origin: http://192.168.1.164:8989
Referer: http://192.168.1.164:8989/
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_14_4) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/73.0.3683.103 Safari/537.36

The related nginx config file.

# Simple http proxy
server 
 listen 8989;
 server_name localhost; 

 # Static grunt server
 location / 
 # Use grunt server (require grunt serve command)
 proxy_pass http://localhost:9001;

 # OR serve static files directly :
 # charset utf-8;
 # root /path/to/docdoku-web-front/dist;
 # expires 0d;
 # access_log off;
 

 # Webservices REST
 location /api 
 proxy_pass http://localhost:8080;
 

 # Websocket application
 location /ws 
 proxy_pass http://localhost:8080/ws;
 proxy_http_version 1.1;
 proxy_set_header Upgrade $http_upgrade;
 proxy_set_header Connection "upgrade"; 
 proxy_read_timeout 7200s;
 

Since it's a separated back-end and front-end project, and can be visit success locally.

So the back-end is correctly config.
The failed request

Request URL: http://localhost:8080/eplmp-server-rest/api/auth/providers

So it's not looking into the right domain. Since the machine A's IP is 192.168.1.164, so the right request URL domain should be there.

It's where the front-end should request, and it's not related with the nginx config file.

Finally, find out there is a config file for the front-end, there's a part config the server domain, which was localhost. After change it to 192.168.1.164 it worked.

No matter what it is, any separated front-end and back-end project should have somewhere for front-end to connect the back-end.

My wonder was since the front-end was deployed on the same server of the back-end, the localhost should work. After I inspect some website, find most of them use a public domain name, like somewebsite.com rather than localhost, though they may not deployed on the same server, this inspired me. Guess there should some kind of front-end client, maybe from the source of the inspect , which need to connect the real back-end.

Really lack of the front-end knowledge, correction is welcome if there are any mistake.