Http_x_forwarded_proto not used by all requests

Hi All,

I have Ghost serving http requests behind Nginx serving SSL.
When i access the site using https it loads, but Chrome complains about some requests:

The page at ‘ - xxx sex videos free hd porn Resources and Information.’ was loaded over HTTPS, but requested an insecure XMLHttpRequest endpoint ‘ - xxx sex videos free hd porn Resources and Information.’. This content should also be served over HTTPS.

Then it tries to load it over http and it fails (http port is not open).

Here are relevant Nginx log entry:
{ “@timestamp”: “2018-08-10T01:29:52+00:00”, “clientip”: “xx.xx.xx.x”, “forwarded_for”: “, xx.xx.xx.x”, “bytes”: 1028, “duration”: 0.001, “status”: 200, “request”: “/public/ghost-sdk.min.js?v=c068d55aa1”, “urlpath”: “/public/ghost-sdk.min.js”, “urlquery”: “v=c068d55aa1”, “method”: “GET”, “upstream-status”: “200”, “http_x_forwarded_host”: “”, “http_x_forwarded_proto”: “https”}

So it looks like http_x_forwarded_proto is propagated, but not used by all requests. How can i fix it?

It’s difficult to determine where the issue is without seeing your site, but based on what you described, the most likely causes are:

  • URL is hardcoded in theme (i.e. <script src=""></script>)
  • Ghost URL is configured for HTTP (i.e. ghost config get url returns

Thanks for reply!

The site is

Your issue is bullet point #2 (Ghost URL is http) -

<link rel="canonical" href="" />

Is the line in the compiled page which shows you this.

If you set up your blog using the Ghost CLI, just run ghost config set url && ghost restart in your installation directory, and everything will work again!

Thanks again!

One complication - I used docker image to deploy it. Is there any way to deal with this?

I’m not exactly sure but something along the lines of redeploy it with the updated URL env variable

bitnami/ghost image has this setting documented as ‘GHOST_HOST’, but I’m using the official image and there is no documentation for this :frowning:

I used the ‘url’ env variable - it seems to have an effect, but now I’m getting ERR_TOO_MANY_REDIRECTS error in Chrome.

The most often cause I’ve seen for this issue is incorrect proxying - can you compare your nginx config to the CLI templates and make changes as needed?

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.