Here is a strange one for you. We have a server with multiple VHOSTS that include both SSL and Non-SSL domains.

Domain1 is SSL enabled, while Domain2 doesn't have SSL.

Since all these domains are hosted on the same IP, apache would respond to httpS requests on domain2 by loading the first SSL enabled vhost, so basically if you went to httpS :// domain2 browser would warn you against an SSL mismatch which would require a user to click on advanced settings prior to seeing the content of Domain1 (the first SSL domain of Apache)

1) If chrome is smart enough to understand there is SSL mismatch, why the heck would Google still index content of Domain1 under https://domain2.com

2) We have since fixed the issue by a putting a re-write that shows 404 for all pages of httpS // domain2.com, We have also used Google Webmaster tools to remove all entries of httpS // domain2.com, however, these keep coming back every 4-6 weeks! I went as far as using Google's fetch URL tool to make sure httpS // domain2.com results into 404 from their point of view and indeed it does.

How the heck is Google still finding content of Domain1 under httpS // Domain2.com? Are they relying on Caches even after a removal request?

All I can think is that Google has the content cached locally and they keep using that content to create indexes again; meaning once we manually request removal of content, they do not crawl the site to re-create that index but they rely on their own local cached copy.

  • It takes some time. Google may have other priorities than fulfiling your request on a single domain. Meanwhile, you could adjust your server a little. 404 isn't probably a best solution; you could use 302 moved permanently to show Google that the canonical address should be http. But realize that Google uses HTTPS as a ranking signal. – Esa Jokinen May 25 '17 at 17:50
  • We had a 302 redirect on https to http, while that was working in browser but the Google index kept coming back, hence why we changed to 404 as a failsafe. – mamad May 25 '17 at 17:53
  • Now it may remove also http results because it doesn't realize that there might be different content on the other protocol, but probably that doesn't matter to you – Esa Jokinen May 25 '17 at 17:56
  • 1
    You could also get a certificate for every domain, implement SNI and abandon HTTP. That would be what professionals do. – Esa Jokinen May 25 '17 at 17:58
  • 1
    302 redirects are not useful for updating search engines (use 301s to do that). Also, it sounds like you may need to implement a default virtualhost (ie. the first VirtualHost encountered on that port.) use httpd -S to find out which one Apache will use. Also, if this is Apache httpd 2.2, make sure you have NameVirtualHost :443 – Cameron Kerr May 26 '17 at 7:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.