502 Bad Gateway (header too large) on Elastic Beanstalk Nginx with Rails 5 App in production

I am using a gatekeeper and building a Rails 5 app to implement my own OAuth provider to connect an Amazon Alexa account. The flow works fine if the user starts the OAuth process and is already registered with my site. But when the user is not signed, he needs to login first and then start the stream. After logging in, the user is not redirected to the OAuth flow. I have now added functionality to redirect back to the OAuth stream after login and it works fine in development mode (using ngrok), but in production I get a 502 error after login and I can't figure out what the problem is ...

Here's an entry from the access.log:

2017/05/03 22:54:04 [error] 2866#0: *185513 upstream sent too big header while reading response header from upstream, client: 172.31.23.232, server: _, request: "GET /oauth/authorize?client_id=08e9435534209c8ee4289ea2bec61a811645b254b060909d142ea3f1f5141600&response_type=token&state=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&redirect_uri=https%3A%2F%2Fpitangui.amazon.com%2Fspa%2Fskill%2Faccount-linking-status.html%3FvendorId%3DM3AP36Y8XL4DFZ HTTP/1.1", upstream: "http://unix:///var/run/puma/my_app.sock:/oauth/authorize?client_id=08e9435534209c8ee4289ea2bec61a811645b254b060909d142ea3f1f5141600&response_type=token&state=eyJpbml0VmVjdG9yIjoiaXRLcmVuWfpbSENBvWfad2VUYW0zZz09IiwicGF5bG9hZCI6IjJOeGU3Rks1TFg2cmJWL2lTM3ZpVXFydTVIdFpiOHM1bGDNbB1wI3M2QkpOU0tKNFVEaXdIdkZnSTBN

      

        
        
        
      

    

(obfuscation values ​​but the length is the same)

and from error.log:

2017/05/04 21:15:07 [error] 579#0: *204674 upstream sent too big header while reading response header from upstream, client: 172.31.23.232, server: _, request: "POST /users/sign_in HTTP/1.1", upstream: "http://unix:///var/run/puma/my_app.sock:/users/sign_in", host: "example.com", referrer: "https://example.com/users/sign_in"

      

        
        
        
      

    

As you can see, the request is very long, so I tried to tweak it large_client_header_buffers

, which had no effect. Then, after reading a bit about similar issues, I tried setting up fastcgi_buffers

and fastcgi_buffer_size

as well as proxy buffers and also no effect. I'm using .ebextension files to add these configurations as described here , but I haven't really found a way to check if these configurations actually took effect after deployment.

Below are the modifications I made to trigger / create before the first error:

Saving back to path in model in block resource_owner_authenticator

:

  resource_owner_authenticator do
    account_link = AccountLink.create(return_to: request.fullpath)
    session[:return_to] = account_link.id

    current_user || warden.authenticate!(:scope => :user)
  end

      

        
        
        
      

    

Redirect to saved path to method after_sign_in_path_for

, if any:

  def after_sign_in_path_for(resource)
    account_link_id = session[:return_to]

    if account_link_id
      account_link = AccountLink.find(account_link_id)

      if account_link
        session.delete(:return_to)

        account_link.return_to
      else
        dashboard_path
      end
    end
  end

      

        
        
        
      

    

I have also seen suggestions that 502s in Elastic Beanstalk might be associated with SSL certificates that do not activate, but I have verified that.

Edit: In my rails handling logs, I see that the message on the path to the sign, as well as the 302 redirect after that was successful. But my browser indicates that the message on the way to the sign has already received 502. I'm not sure what to do with that.