Upstream prematurely closed FastCGI stdout while reading response header from upstream - php

Question

I suddently get a 502 Bad gateway error and I don't understand why this error appear. This error appear moreover only for 1 single page!!...
The exact error in my Nginx log is :
Upstream prematurely closed FastCGI stdout while reading response header from upstream [..] upstream: "fastcgi://unix:/var/run/php5-fpm.sock:"
I tried :
service nginx restart : NOTHING CHANGE
service php5-fpm restart : NOTHING CHANGE
to reboot the server : NOTHING CHANGE
even to restart the mysql service : NOTHING CHANGE
My /var/log/upstart/php5-fpm.log (only a lot of NOTICES) :
Terminating...
exiting, bye-bye!
fpm is running, pid 9887
ready to handle connections
systemd monitor interval set to 10000ms
that makes me crazy, any idea ?

Answer 1

Strangely enough, this is what worked for me after chasing down a mysterious 502 Bad Gateway error:
Change error_reporting setting in php-fpm ini like this:
error_reporting = ~E_ALL
Then restart php-fpm and nginx
This worked for me after trying the suggested nginx settings above to no avail. I don't have a verified explanation for this, but it would appear that excessive error reporting can overload the php-fpm process.

Answer 2

Try to include <fcgi_stdio.h> in your source file.
This error happens to me just now,then i add include <fcgi_stdio.h>in my c file,here is my code.
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <fcgi_config.h>
#include <fcgi_stdio.h>
int main()
{
int count = 0;
while(FCGI_Accept() >= 0)
{
printf("content-type:text/html\r\n");
printf("\r\n");
printf("<title>Fast CGI Hello</title>");
printf("<h1>fast CGI hello</h1>");
printf("Request number %d running on host<i>%s</i>\n",++count,getenv("SERVER_NAME"));
}
return 0;
}

Answer 3

I had the error only for one type of requests that had 12 items in the body. May be it was related with a body size, but it worked when it was more (13 and more items) or some less (11 and less).
upstream prematurely closed FastCGI stdout while reading response header from upstream
2021/12/24 07:14:44 [error] 9#9: *49 upstream prematurely closed FastCGI stdout while reading response header from upstream, client: 172.21.0.1, server: , request: "PUT /v1/MY_PAGE/a976d2e5-afc0-4f6d-8d05-17af7dc73f46 HTTP/1.1", upstream: "fastcgi://172.21.0.3:9000", host: "localhost"
I used Docker containers with Nginx and PHP-FPM.
Nginx returned 502 Bad Gateway.
PHP-FPM was used XDebug and it made the error.
When I turned on debugging the error left and when I turned off debugging the error returned.
Solution
To solve the error I disabled XDebug. Also it works when I removed XDebug.
Options to solve the error (one of these):
Disable XDebug (one of these):
Set the XDEBUG_MODE='off' environment variable.
Set xdebug.mode=off in xdebug.ini or another PHP config.
https://xdebug.org/docs/all_settings#mode
Remove XDebug from your PHP server and restart PHP-FPM (one of these):
Remove through PECL (works with Docker): pecl uninstall xdebug.
Remove through Ubuntu APT: sudo apt-get purge php-xdebug.
Restart your PHP-FPM or your service/server.

Answer 4

I did the following to server nginx config:
client_body_timeout 1200;
client_header_timeout 600;
and added zend_extension to php fpm php.ini:
zend_extension = xdebug.so