If nginx is running as a proxy / reverse proxy

that is, for users of ngx_http_proxy_module

In addition to fastcgi, the proxy module also saves the request header in a temporary buffer.

So you may need also to increase the proxy_buffer_size and the proxy_buffers, or disable it totally (Please read the nginx documentation).

Example of proxy buffering configuration

http { proxy_buffer_size 128k; proxy_buffers 4 256k; proxy_busy_buffers_size 256k; } 

Example of disabling your proxy buffer (recommended for long polling servers)

http { proxy_buffering off; } 

For more information: Nginx proxy module documentation

Plesk instructions

I combined the top two answers here

In Plesk 12, I had nginx running as a reverse proxy (which I think is the default). So the current top answer doesn't work as nginx is also being run as a proxy.

I went to Subscriptions | [subscription domain] | Websites & Domains (tab) | [Virtual Host domain] | Web Server Settings.

Then at the bottom of that page you can set the Additional nginx directives which I set to be a combination of the top two answers here:

fastcgi_buffers 16 16k; fastcgi_buffer_size 32k; proxy_buffer_size 128k; proxy_buffers 4 256k; proxy_busy_buffers_size 256k; 

upstream sent too big header while reading response header from upstream is nginx's generic way of saying "I don't like what I'm seeing"

1. Your upstream server thread crashed
2. The upstream server sent an invalid header back
3. The Notice/Warnings sent back from STDERR overflowed their buffer and both it and STDOUT were closed

3: Look at the error logs above the message, is it streaming with logged lines preceding the message? PHP message: PHP Notice: Undefined index: Example snippet from a loop my log file:

2015/11/23 10:30:02 [error] 32451#0: *580927 FastCGI sent in stderr: "PHP message: PHP Notice: Undefined index: Firstname in /srv/www/classes/data_convert.php on line 1090 PHP message: PHP Notice: Undefined index: Lastname in /srv/www/classes/data_convert.php on line 1090 ... // 20 lines of same PHP message: PHP Notice: Undefined index: Firstname in /srv/www/classes/data_convert.php on line 1090 PHP message: PHP Notice: Undefined index: Lastname in /srv/www/classes/data_convert.php on line 1090 PHP message: PHP Notice: Undef 2015/11/23 10:30:02 [error] 32451#0: *580927 FastCGI sent in stderr: "ta_convert.php on line 1090 PHP message: PHP Notice: Undefined index: Firstname 

you can see in the 3rd line from the bottom that the buffer limit was hit, broke, and the next thread wrote in over it. Nginx then closed the connection and returned 502 to the client.

2: log all the headers sent per request, review them and make sure they conform to standards (nginx does not permit anything older than 24 hours to delete/expire a cookie, sending invalid content length because error messages were buffered before the content counted...). getallheaders function call can usually help out in abstracted code situations php get all headers

examples include:

<?php //expire cookie setcookie ( 'bookmark', '', strtotime('2012-01-01 00:00:00') ); // nginx will refuse this header response, too far past to accept .... ?> 

and this:

<?php header('Content-type: image/jpg'); ?> <?php //a space was injected into the output above this line header('Content-length: ' . filesize('image.jpg') ); echo file_get_contents('image.jpg'); // error! the response is now 1-byte longer than header!! ?> 

1: verify, or make a script log, to ensure your thread is reaching the correct end point and not exiting before completion.

Add:

fastcgi_buffers 16 16k; fastcgi_buffer_size 32k; proxy_buffer_size 128k; proxy_buffers 4 256k; proxy_busy_buffers_size 256k; 

To server{} in nginx.conf

Works for me.

I have a django application deployed to EBS, and I am using Python 3.8 running on 64bit Amazon Linux 2. The following method worked for me (note folder structure might be DIFFERENT if you're using previous Linux versions. For more, see official documentation here

Make the .platform folder and its sub-directory as shown below:

|-- .ebextensions # Don't put nginx config here | |-- django.config |-- .platform # Make ".platform" folder and its subfolders |-- nginx | -- conf.d | -- proxy.conf 

Note that proxy.conf file should be placed inside .platform folder, NOT .ebextensions folder or the .elasticbeanstalk folder. The extension should end with .conf NOT .config.

Inside the proxy.conf file, copy & paste these lines directly:

client_max_body_size 50M; large_client_header_buffers 4 32k; fastcgi_buffers 16 32k; fastcgi_buffer_size 32k; proxy_buffer_size 128k; proxy_buffers 4 256k; proxy_busy_buffers_size 256k; 

There is no need to issue command to restart nginx (for Amazon Linux 2)

Deploy the source code to elastic beanstalk again.

Faced the same problem when running Symfony app in php-fpm and nginx in docker containers.

After some research found that it was caused by php-fpm's stderr written to nginx logs. I.e. php warnings (which is generated intensively in Symfony debug mode) was duplicated in docker logs php-fpm:

[09-Jul-2021 12:25:46] WARNING: [pool www] child 38 said into stderr: "" [09-Jul-2021 12:25:46] WARNING: [pool www] child 38 said into stderr: "NOTICE: PHP message: [debug] Notified event "kernel.response" to listener "Symfony\Component\HttpKernel\EventListener\DisallowRobotsIndexingListener::onResponse"." [09-Jul-2021 12:25:46] WARNING: [pool www] child 38 said into stderr: "" [09-Jul-2021 12:25:46] WARNING: [pool www] child 38 said into stderr: "NOTICE: PHP message: [debug] Notified event "kernel.response" to listener "Symfony\Component\HttpKernel\EventListener\StreamedResponseListener::onKernelResponse"." [09-Jul-2021 12:25:46] WARNING: [pool www] child 38 said into stderr: "" [09-Jul-2021 12:25:46] WARNING: [pool www] child 38 said into stderr: "NOTICE: PHP message: [debug] Notified event "kernel.finish_request" to listener "Symfony\Component\HttpKernel\EventListener\LocaleListener::onKernelFinishRequest"." [09-Jul-2021 12:25:46] WARNING: [pool www] child 38 said into stderr: "" [09-Jul-2021 12:25:46] WARNING: [pool www] child 38 said into stderr: "NOTICE: PHP message: [debug] Notified event "kernel.finish_request" to listener "Symfony\Component\HttpKernel\EventListener\RouterListener::onKernelFinishRequest"." [09-Jul-2021 12:25:46] WARNING: [pool www] child 38 said into stderr: "" [09-Jul-2021 12:25:46] WARNING: [pool www] child 38 said into stderr: "NOTICE: PHP message: [debug] Notified event "kernel.finish_request" to listener "Symfony\Component\HttpKernel\EventListener\LocaleAwareListener::onKernelFinishRequest"." [09-Jul-2021 12:25:46] WARNING: [pool www] child 38 said into stderr: "" [09-Jul-2021 12:25:46] WARNING: [pool www] child 38 said into stderr: "NOTICE: PHP message: [debug] Notified event "kernel.terminate" to listener "Symfony\Component\HttpKernel\EventListener\ProfilerListener::onKernelTerminate"." [09-Jul-2021 12:25:46] WARNING: [pool www] child 38 said into stderr: "" 

and docker logs nginx:

2021/07/09 12:25:46 [error] 30#30: *2 FastCGI sent in stderr: "ller" to listener "OblgazAPI\API\Common\Infrastructure\EventSubscriber\LegalAuthenticationChecker::checkAuthentication". PHP message: [debug] Notified event "kernel.controller_arguments" to listener "Symfony\Component\HttpKernel\EventListener\ErrorListener::onControllerArguments". PHP message: [debug] Notified event "kernel.response" to listener "Symfony\Component\HttpKernel\EventListener\ResponseListener::onKernelResponse". PHP message: [debug] Notified event "kernel.response" to listener "Symfony\Component\HttpKernel\DataCollector\RequestDataCollector::onKernelResponse". PHP message: [debug] Notified event "kernel.response" to listener "Symfony\Component\HttpKernel\EventListener\ProfilerListener::onKernelResponse". PHP message: [debug] Notified event "kernel.response" to listener "Symfony\Bundle\WebProfilerBundle\EventListener\WebDebugToolbarListener::onKernelResponse". PHP message: [debug] Notified event "kernel.response" to listener "Symfony\Component\HttpKernel\EventListener\DisallowRobotsIndexingListener::onResponse". PHP message: [debug] Notified event "kernel.response" to listener "Symfony\Component\HttpKernel\EventListener\StreamedResponseListener::onKernelResponse". PHP message: [debug] Notified event "kernel.finish_request" to listener "Symfony\Component\HttpKernel\EventListener\LocaleListener::onKernelFinishRequest". PHP message: [debug] Notified event "kernel.finish_request" to listener "Symfony\Component\HttpKernel\EventListener\RouterListener::onKernelFinishRequest". PHP message: [debug] Notified event "kernel.finish_request" to listener "Symfony\Component\HttpKernel\EventListener\LocaleAwareListener::onKernelFinishRequest". PHP message: [debug] Notified event "kernel.exception" to listener "Symfony\Component\HttpKernel\EventListener\ErrorListener::logKernelException". PHP message: [debug] Notified event "kernel.exception" to listener "Symfony\Component\HttpKernel\EventListener\ProfilerListener::onKernelException". 

and then nginx logs ended with

2021/07/09 12:25:46 [error] 30#30: *2 upstream sent too big header while reading response header from upstream ... 

and I got 502 error.

Increasing the fastcgi_buffer_size in nginx config helped, but it seems more like a suppression the problem, not a treatment.

A better solution is to disable php-fpm to send logs by FastCGI. Found it can be made by setting fastcgi.logging=0 in php.ini (by default it is 1). php docs.

After changing it to 0, the problem goes away and nginx logs looks much cleaner docker logs nginx:

172.18.0.1 - - [09/Jul/2021:12:36:02 +0300] "GET /my/symfony/app HTTP/1.1" 401 73 "-" "PostmanRuntime/7.26.8" 172.18.0.1 - - [09/Jul/2021:12:36:04 +0300] "GET /my/symfony/app HTTP/1.1" 401 73 "-" "PostmanRuntime/7.26.8" 

and all php-fpm logs are still in their place in php-fpm log.

fastcgi_busy_buffers_size 512k; fastcgi_buffer_size 512k; fastcgi_buffers 16 512k; 

it worked for me when I increased the numbers

We ended up realising that our one server that was experiencing this had busted fpm config resulting in php errors/warnings/notices that'd normally be logged to disk were being sent over the FCGI socket. It looks like there's a parsing bug when part of the header gets split across the buffer chunks.

So setting php_admin_value[error_log] to something actually writeable and restarting php-fpm was enough to fix the problem.

We could reproduce the problem with a smaller script:

<?php for ($i = 0; $i<$_GET['iterations']; $i++) error_log(str_pad("a", $_GET['size'], "a")); echo "got here\n"; 

Raising the buffers made the 502s harder to hit but not impossible, e.g native:

bash-4.1# for it in {30..200..3}; do for size in {100..250..3}; do echo "size=$size iterations=$it $(curl -sv "http://localhost/debug.php?size=$size&iterations=$it" 2>&1 | egrep '^< HTTP')"; done; done | grep 502 | head size=121 iterations=30 < HTTP/1.1 502 Bad Gateway size=109 iterations=33 < HTTP/1.1 502 Bad Gateway size=232 iterations=33 < HTTP/1.1 502 Bad Gateway size=241 iterations=48 < HTTP/1.1 502 Bad Gateway size=145 iterations=51 < HTTP/1.1 502 Bad Gateway size=226 iterations=51 < HTTP/1.1 502 Bad Gateway size=190 iterations=60 < HTTP/1.1 502 Bad Gateway size=115 iterations=63 < HTTP/1.1 502 Bad Gateway size=109 iterations=66 < HTTP/1.1 502 Bad Gateway size=163 iterations=69 < HTTP/1.1 502 Bad Gateway [... there would be more here, but I piped through head ...] 

fastcgi_buffers 16 16k; fastcgi_buffer_size 32k;:

bash-4.1# for it in {30..200..3}; do for size in {100..250..3}; do echo "size=$size iterations=$it $(curl -sv "http://localhost/debug.php?size=$size&iterations=$it" 2>&1 | egrep '^< HTTP')"; done; done | grep 502 | head size=223 iterations=69 < HTTP/1.1 502 Bad Gateway size=184 iterations=165 < HTTP/1.1 502 Bad Gateway size=151 iterations=198 < HTTP/1.1 502 Bad Gateway 

So I believe the correct answer is: fix your fpm config so it logs errors to disk.

If you're using Symfony framework: Before messing with Nginx config, try to disable ChromePHP first.

1 - Open app/config/config_dev.yml

2 - Comment these lines:

#chromephp: #type: chromephp #level: info 

ChromePHP pack the debug info json-encoded in the X-ChromePhp-Data header, which is too big for the default config of nginx with fastcgi.

Source: https://github.com/symfony/symfony/issues/8413#issuecomment-20412848

There are two cases in which I have detect that this happens.

1.- Large data, this is resolved with the accepted answer

1.1.- Setting buffer size

 #nginx proxy_busy_buffers_size 1024k; proxy_buffers 32 1024k; proxy_buffer_size 1024k; #fastcgi fastcgi_buffers 32 1024k; fastcgi_buffer_size 1024k; fastcgi_busy_buffers_size 1024k; 

1.2.- Or turn off buffers

 #nginx proxy_buffering off; #fastcgi fastcgi_buffering off; 

(como dice Meo, "Lo que te flote el barco")

2.- Trying to respond to binary data as text

if you are responding to a JSON and you data contains some binary, you get the same error. For example, something like this:

echo json_encode($data) 

then if you want to respond to a json that contains some binary data, you must first perform a deep conversion like:

$data = UtilBin::convertBinToBase64($data); echo json_encode($data) 

here my simple convert, that check if any string field is binary

class UtilBin{ public static function convertBinToBase64($data) { if($data === null){ return null; } if (self::isBinaryData($data)) { return base64_encode($data); } if(is_resource($data)){ $data = stream_get_contents($data); return base64_encode($data); } if (is_array($data) || is_object($data)) { foreach ($data as $key => $value) { $new_val = $value; if (is_array($value) || is_object($value)) { $new_val = self::convertBinToBase64($value); // Recursive call for nested arrays or objects } elseif (self::isBinaryData($value)) { // Check if $value is binary $new_val = base64_encode($value); } if(is_array($data)){ $data[$key] = $new_val; }elseif(is_object($data)){ $data->$key = $new_val; } } } return $data; } private static function isBinaryData($str) { if($str === null){ return false; } if(!is_string($str)){ return false; } // Use a regular expression to check for non-printable ASCII characters return preg_match('/[^\x20-\x7E]/', $str) > 0; } } 

I had this error and I found 3 ways to fix that:

• Set SHELL_VERBOSITY=0 or another value < 3 in .env: https://stackoverflow.com/a/69321273/13653732 In this case you disable your PHP logs, but they can be useful for development and debugging.
• Set fastcgi.logging=0 in php.ini. It is the same result like above.
• Update Symfony from 5.2 to 5.3. I think the old version has a problem with that.

All PHP Symfony logs perceived like errors for Nginx, but PHP worked correctly.

I had Nginx 1.17, PHP 8.0.2, PHP-FPM, Symfony 5.2, Xdebug, Docker.

I tried new versions Nginx 1.21, PHP 8.0.14, but without any result. That problem wasn't with Apache.

I changed Nginx configuration, but also without any result.

This is still the highest SO-question on Google when searching for this error, so let's bump it.

When getting this error and not wanting to deep-dive into the NGINX settings immediately, you might want to check your outputs to the debug console. In my case I was outputting loads of text to the FirePHP / Chromelogger console, and since this is all sent as a header, it was causing the overflow.

It might not be needed to change the webserver settings if this error is caused by just sending insane amounts of log messages.

I am not sure that the issue is related to what header php is sending. Make sure that the buffering is enabled. The simple way is to create a proxy.conf file:

proxy_redirect off; proxy_set_header Host $host; proxy_set_header X-Real-IP $remote_addr; proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; client_max_body_size 100m; client_body_buffer_size 128k; proxy_connect_timeout 90; proxy_send_timeout 90; proxy_read_timeout 90; proxy_buffering on; proxy_buffer_size 128k; proxy_buffers 4 256k; proxy_busy_buffers_size 256k; 

And a fascgi.conf file:

fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name; fastcgi_param QUERY_STRING $query_string; fastcgi_param REQUEST_METHOD $request_method; fastcgi_param CONTENT_TYPE $content_type; fastcgi_param CONTENT_LENGTH $content_length; fastcgi_param SCRIPT_NAME $fastcgi_script_name; fastcgi_param REQUEST_URI $request_uri; fastcgi_param DOCUMENT_URI $document_uri; fastcgi_param DOCUMENT_ROOT $document_root; fastcgi_param SERVER_PROTOCOL $server_protocol; fastcgi_param GATEWAY_INTERFACE CGI/1.1; fastcgi_param SERVER_SOFTWARE nginx/$nginx_version; fastcgi_param REMOTE_ADDR $remote_addr; fastcgi_param REMOTE_PORT $remote_port; fastcgi_param SERVER_ADDR $server_addr; fastcgi_param SERVER_PORT $server_port; fastcgi_param SERVER_NAME $server_name; fastcgi_buffers 128 4096k; fastcgi_buffer_size 4096k; fastcgi_index index.php; fastcgi_param REDIRECT_STATUS 200; 

Next you need to call them in your default config server this way:

http { include /etc/nginx/mime.types; include /etc/nginx/proxy.conf; include /etc/nginx/fastcgi.conf; index index.html index.htm index.php; log_format main '$remote_addr - $remote_user [$time_local] $status ' '"$request" $body_bytes_sent "$http_referer" ' '"$http_user_agent" "$http_x_forwarded_for"'; #access_log /logs/access.log main; sendfile on; tcp_nopush on; # ........ } 

In our case we got this nginx error because our backend generated redirect response with a very long URL:

HTTP/1.1 301 Moved Permanently Location: https://www.example.com/?manyParamsHere... 

Just for curiosity, we saved that big URL to a file and it size was 4.4 Kb.

Adding two lines to the config file /etc/nginx/conf.d/some_site.conf helped us to fix this error:

server { # ... location ~ ^/index\.php(/|$) { fastcgi_pass php-upstream; fastcgi_split_path_info ^(.+\.php)(/.*)$; include fastcgi_params; fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name; # Add these lines: fastcgi_buffer_size 32k; fastcgi_buffers 4 32k; } } 

Read more about these params at the official nginx documentation.

For Symfony projects, try adding this line to your .env file:

SHELL_VERBOSITY=0 

https://symfony.com/doc/current/console/verbosity.html

I will just leave it here, because I spent a crazy amount of time debugging this in my project and only this particular solution worked 100% for me (for the right reasons) and I have never found this answer related to this topic. Maybe someone will find it helpful.

For the ones who face this issue with nginx, Laravel and Vite:

• the upstream header leading to the problem is the Link header being too long
• sure, the server should support it as is, but sometimes you do not have access to change the buffer size from other answers
• Laravel adds this header to make assets loading slightly faster with preloading mechanics

Laravel class AddLinkHeadersForPreloadedAssets is responsible for adding Link header, therefore, remove it or comment in your Kernel.php configuration in the $middlewares section.

 protected $middlewareGroups = [ 'web' => [ \App\Http\Middleware\EncryptCookies::class, //... // \Illuminate\Http\Middleware\AddLinkHeadersForPreloadedAssets::class, \App\Http\Middleware\HandleInertiaRequests::class, ] ] 

I had this error today in Nginx error_log, my Nginx is configured to load data from Apache.

upstream sent too big header while reading response header from upstream 

In my case it was caused by trying to set a cookie in PHP that reached over 4,096 bytes

So first, check if that is the case, before increasing the limits.

For example you can make a cookie huge by saving a user text field in it, or by appending data repeatedly to existent cookie value.

Sending a very large cookie with every request to the site (every image, css, etc) can slow down the site a lot.

Initially I fixed it by adding this to my Nginx server{} context (can be http{}, server{} or location{} context):

proxy_buffer_size 128k; # Sets both the number and size of the buffers used proxy_buffers 8 256k; proxy_busy_buffers_size 256k; 

Nginx manual page.

I am using Symfony, it has a nice exception page (when the ErrorHandler is used). And this one will put the message of your exception as a header to the response created.

vendor/symfony/error-handler/ErrorRenderer/SerializerErrorRenderer.php

the header is called: X-Debug-Exception

So be careful, if you constructed a VERY large exception message, neither nginx nor chrome (limit 256k) nor curl (~128kb) can display your page and make it really hard to debug, what is outputting those big headers.

My suggestion would be to not blindly copy n paste increased buffer sizes into your nginx config, they treat the symptom not the cause.