The company has a front end of the project where the server is rendered using node. Js and then returned to the browser to solve the soo problem on a single page. When the project was deployed, nginx was used as a reverse agent, node. Js. Specific steps are as follows:
(nginx, node. Js installation and basic configuration directly skipping)
First we have to open the following configuration at the http node in the nginx. Cnf file:
Http {
log formatmain '$remote addr - $remote user [$time local] "$request"
'$status $body bytes cent "$http referer"'
'$http user ageThe configuration file for each domain name is then placed under this directory/etc/nginx/conf. D/, with the file suffix ending with a conf。
The first, the simple:
Server {
80;
name localhost;
root /xx/xx/hxydex/;
#set $my server name $scheme://$server name;
#if ($my server name! =https://$server name) {
# rewrite # https://$server name$request uri?
♪ i'm sorry ♪
er
access log/var/log/nginx/hyde accss. Logmain;
organisation
proxy set header x-real-ip $remote addr;
proxy set header x-for$proxy add x forwarded for;
proxy set head host $ http host;
proxy set header x-nginx-proxy true;
proxy http version1. 1;
proxy set head coNo, no, no.
# no need to consider loads, no need to configure upstream nodes. Proxy pass http://127. 0. 0. 1:3000;
♪ i'm sorry ♪
(b) error page 404/404. HTML;
localization = /xx/xx/40x. HTML
♪ i'm sorry ♪
error page 500 502, 503,504/50x. HTML;
localization = /xx/xx/50x. HTML
♪ i'm sorry ♪
♪ i'm sorry ♪
Second approach, taking into account load
I'm sorry, upstream node
server 127. 0. 0. 1:3,000;
♪ i'm sorry ♪
server {
80;
name localhost;
root /xx/xx/hxydex/;
#set $my server name $scheme://$server name;
#if ($my server name! =https://$server name) {
# rewrite # https://$server name$request uri?
♪ i'm sorry ♪
er
access log/var/log/nginx/hyde accss. Logmain;
organisation
proxy set header x-real-ip $remote addr;
proxy set header x-for$proxy add x forwarded for;
proxy set head host $ http host;
proxy set header x-nginx-proxy true;
proxy http version1. 1;
proxy set head coNo, no, no.
# configureupstream node
proxy pass http://node;
♪ i'm sorry ♪
(b) error page 404/404. HTML;
localization = /xx/xx/40x. HTML
♪ i'm sorry ♪
error page 500 502, 503,504/50x. HTML;
localization = /xx/xx/50x. HTML
♪ i'm sorry ♪
♪ i'm sorry ♪
Then you can restart or reload the nginx profile. The order reads as follows:
#check if nginx profile is correct
i don't know.
# restart nginx
i don't know.
# reload profile
i don't know
Questions noted:
The following problems may arise:
It's not like i'm going to be able to do this.
/ unhandled 'error' event
yeah. Error: listen to eaddrinuse
(net. Js: 884:11)
listen2 (net. Js:1022:14)
at listen (net. Js: 1044:10)
at server. Listen (net. Js:1110:5)
i don't know, at object.(foldername/app. Js:33:24)
(module. Js: 456:26)
(module. Js: 474:10)
at modeule. Load (module. Js: 356:32)
(module. Js: 312:12)
at function. Module. Runmain
This is actually an error caused by the occupation of the node. Js service multi-starter, a problem that could arise by using the node. Js project management tool pm2 or using the netstat-anop to view the port occupied by that process and then kill the restart service
Add a load balance strategy for nginx:
Rotation (default)
Each request is assigned in chronological order to a different backend server, which can be automatically eliminated if the backend server down。
I'm sorry, upstream backserver
server 192. 168. 0. 14;
server 192. 168. 0. 15;
♪ i'm sorry ♪
Assign weight
Specifies the probability of a rotation, which is proportional to the ratio of access, for the back-end server performance。
I'm sorry, upstream backserver
server 192. 168. 0. 14 weight = 10;
server 192. 16. 8. 15 weight = 10;
♪ i'm sorry ♪
Ip binding ip hash
Each request is distributed according to the hash results of the visit to ip, so that each visitor has regular access to a back-end server and can solve the problem of the session。
I'm sorry, upstream backserver
ip hash;
server 192. 168. 0. 14:88;
server 192. 16. 8. 15:80;
♪ i'm sorry ♪
Fair (third party)
Distribution of requests according to back-end server response time and response to short priority assignments。
I'm sorry, upstream backserver
server 192. 168. 0. 14:88;
server 192. 16. 8. 15:80;
i'm sorry.
♪ i'm sorry ♪
Url hash (third party)
Distributes requests according to the rash results of access to url, directing each url to the same back-end server, which is more effective when cached。
I'm sorry, upstream backserver
server squid 1:3128;
server squid 2:3128;
hash $request uri;
hash method crc32;
♪ i'm sorry ♪
Summary
This is the full text of the article, which i hope will be of some interest to you in your studies or your work. If you have any questions, please leave a message and thank you for your support。
