413 Request Entity Too Large

[tigpt]

Error description:

in some php forums like SMF, some requests like uploading files I get 413 Request Entity Too Large

Steps to reproduce

Use large urls

In conf.d/uploadsize.conf I change the value of client_max_body_size 100m; to client_max_body_size 100M; or even to 0; to disable it and I keep getting the same error?

Could the error came from the nginx behind the proxy and the proxy just relays it? changing the value dissent differs anything.

I get the error from the proxy since is the one running nginx 1.15.2, the nginx serving the files is on 1.14 and error page show 1.15.2

Is the client_max_body_size being loaded?

[evertramos]

After you change the value did you restart the service?

$ docker exec -it nginx-web nginx -s reload

Your backend container which passed through the proxy should have a valid configuration in order to accept the file upload size.

[tigpt]

No, but I did docket-compose down and docker-compose up -d

Everything stayed the same.

I tried docker-compose down and the
docker system prune
Then ./start.sh to pull and start again

It still says the same error when I try to upload a file

Edit: now I've done docker exec -it nginx-web nginx -s reload and still the same result.

how can I debug if this setting is being loaded to nginx?

[tigpt]

docker exec -it nginx-web nginx -T doesn't show me client_max_body_size, is it suppose to?

found this:

client_max_body_size can be set on 3 levels: global, virtualhost, or location. This is something that should be in your nginx virualhost settings.

http://nginx.org/en/docs/http/ngx_http_core_module.html#client_max_body_size
Show that context can be

Context: http, server, location

just for be sure its not something on the website config, I changed the server config also to the same value, even not believing it was the problem, since the error says nginx 1.15.2 that is only running on the proxy.

Everything looks the same.

[evertramos]

Did you enabled it on the .env file?

Uncomment this line in your ".env" file:

USE_NGINX_CONF_FILES=true

And do a compose down and up again.

[tigpt]

(>ლ)
You are so patient with my issues, god bless you.

Thanks a lot, you are a life saver.
After uncomment USE_NGINX_CONF_FILES=true and relaunch with ./start.sh it was working.

(well, it started showing the same but for 1.14 (nginx of the site, not the proxy), and then it needed to change the php.ini, but it finally worked as expected)

[evertramos]

Well... if the issue is in your nginx... you add this line to your nginx conf:

https://github.com/evertramos/nignx-behind-proxy/blob/master/nignx/conf.d/default.conf#L19

Hope it helps!

God bless you too!

Thanks for the kind words.

[tionebDotsamapha]

Did you enabled it on the .env file?

Uncomment this line in your ".env" file:

USE_NGINX_CONF_FILES=true

And do a compose down and up again.

Thanks guys! A great help for me!

[tionebDotsamapha]

Finally that not work for me...

To soluce the problem I modified the nginx.tmpl to add client_max_body_size 100m; in all location /{}

I don't know if it's a good practice but it's works

[evertramos]

There is not problem to add that into your location! =)

Just make sure you really need 100 mb of data in your body for all of your locations.

[hayzem]

FYI, creating a config file under conf.d is synchronized with nginx-data directory. But after I removed some config files it kept the old ones in nginx-data and kept giving me this error.

It wold be nice to add functionality to sync properly or deleting nginx-data with a flag like
./start.sh --clean

[ingen2]

Please add environment variable (per docker-compose project) for setting client_max_body_size;

docker-compose.yml file Example:

   nginx:
   build: docker/nginx
   container_name: "myapp_nginx"
   environment:
     CLIENT_NAX_BODY_SIZE: "100m"
     VIRTUAL_HOST: "didarnet.ir"
     LETSENCRYPT_HOST: "didarnet.ir"
     LETSENCRYPT_EMAIL: "admin@didarnet.ir"

`