16

I don't know why I got this error every time I tried to open the page:

2013/04/06 17:52:19 [error] 5040#0: *1 connect() failed (111: Connection refused) while connecting to upstream, client: 127.0.0.1, server: localhost, request: "GET /info.php HTTP/1.1", upstream: "fastcgi://127.0.0.1:9000", host: "localhost:8080"
Hazem Hagrass
  • 9,329
  • 10
  • 32
  • 54

5 Answers5

39

I resolved it, it was a configuration file issue, I added:

location ~ .php$ {
    try_files $uri =404;
    fastcgi_split_path_info ^(.+\.php)(/.+)$;
    fastcgi_pass unix:/var/run/php5-fpm.sock;
    fastcgi_index index.php;
    include fastcgi_params;
}
Limon Monte
  • 52,539
  • 45
  • 182
  • 213
Hazem Hagrass
  • 9,329
  • 10
  • 32
  • 54
  • 4
    There are some situations when `fastcgi_pass 127.0.0.1:9000;` is present in server block and is responsible for this error, in such cases you need change it to `fastcgi_pass unix:/var/run/php5-fpm.sock;`. – xyz Dec 08 '13 at 10:34
  • 2
    This is because php5-fpm now ships with the Unix socket `listen = /var/run/php5-fpm.sock` enabled in `/etc/php5/fpm/pool.d/www.conf` instead of the TCP socket `listen = 127.0.0.1:9000`. This is typically a little faster. (For google search purposes) `ubuntu 14.04 nginx php5-fpm 502 Bad Gateway`. I have a feeling a lot of people will be needing this in the near future. There are so many tutorials with the old socket in place. – DutGRIFF May 09 '14 at 06:09
  • Thank you, @KamilZieliński! Changing to `fastcgi_pass unix:/var/run/php5-fpm.sock;` fixed my error. – Ryan Jun 04 '14 at 18:31
  • In my case (centos 7) I had to use `fastcgi_pass unix:/var/run/php-fpm/php-fpm.sock;`. The error log is your friend:`sudo tail -20 /var/log/nginx/error.log` – Peter Smartt Feb 27 '15 at 04:34
18

For me the problem was my php-fpm service wasn't running. You can check it by running:

service php-fpm status

and start it by running

service php-fpm start

Sometimes php-fpm might have broken instances running, preventing a restart. This command is a clean way to clear them out and restart php-fpm

killall -9 php-fpm; service php-fpm restart
Syntax Error
  • 4,475
  • 2
  • 22
  • 33
ContextSwitch
  • 2,830
  • 6
  • 35
  • 51
3

update your configurations as mentioned before:

location ~ .php$ {
try_files $uri =404;
fastcgi_split_path_info ^(.+\.php)(/.+)$;
fastcgi_pass unix:/var/run/php5-fpm.sock;
fastcgi_index index.php;
include fastcgi_params;
}

but don't forget to restart both nginx server and php-fpm after updating

sudo /etc/init.d/nginx restart
sudo /etc/init.d/php-fpm restart
user3111020
  • 66
  • 2
  • This belongs in a comment below the answer given. Kinda looks like you are just trying to get upvotes. Yes. Restarting nginx is necessary but most will know this and if they don't they will see it in a comment before they scroll all the way down here. – DutGRIFF May 09 '14 at 02:44
3

I found I had this same issue with PHP7 running in Docker on a Debian Jessie (8.3) instance.

  • running command 'ps -aux' showed that php-fpm wasn't running
  • running 'php-fpm -D' brought it up as deamonized process.
  • Rerunning 'ps -aux' showed that php-fpm was indeed running
  • Refreshing my test page showed me the servers PHP info.

Added 'php-fpm -D' to my start.sh script so that things started every time the container is loaded.

Hope this helps someone.

Lionel Morrison
  • 566
  • 4
  • 15
0

Use fastcgi_pass unix:/var/run/php5-fpm.sock; only nginx and php install same server. If nginx and php install in other server you must use fastcgi_pass ip server:port;

Scott Zụ
  • 11
  • 2