3

Back on 2/7, we changed CDNs from Amazon Cloudfront to StackPath for the images on our popular sports website. We ticked off the "use canonical headers" option when doing so. Starting on 2/12, we started to see crazy crawl stats in Google Search Console. On 2/17, our search traffic completely fell off. Starting on 2/17, we have been facing a 4-5 hour delay in our content hitting Google or Google News.

We then realized that between 2/12 and 2/19, our "indexed pages" in Google's search console shot up from 63,000 to 89,000. Over the next seven days, it continued up - to 98,000. I'm convinced that the sudden spike in indexed pages is the reason for the traffic drop and the delay. Between our articles, pages and category pages we should be around that 63,000 number.

When doing a site: search, Google only shows 58,500 pages. Bing shows 57,600. So I don't think we have duplicate content that I'd be able to find in the conventional way.

I've also parsed through the server log files. There aren't any weird URLs with ? type of parameters.

We have 45,000 images on our site. Is it possible that Google in now indexing those images and counting them in its "indexed pages" report? That's the only thing I can think of - but I've heard conflicting information about whether Google counts images in that number.

Improve this question
3
  • Hi mate, have you tried using site: and inurl: when searching for image URLs that have been indexed to confirm is images the issue? It could be that your CDN is serving diferent variations with parameters of the same asset (diferent image size, location, etc)
    – Raul Reyes
    May 7, 2017 at 22:27
  • Had an issue like this. Crawled pages grew & newly indexed pages should of been 404s . What was happening was the CMS was serving "pages" for URL structures that didn't exist and should of been going to 404 pages. The way we resolved this was we used an if statement that would redirect to 404 pages if the page didn't exist. It's hard to see if you are having the same issue because I don't know what your site is and I can't see if unreal URLs works. Try going to URL that you know is a page like: www.example.com/sports/gear and then add /test at the end and see if it takes you to a 404
    – Trent
    May 23, 2017 at 7:25
  • Is setup of image's MIME type correct? Inspect whether your images are delivered as images in terms of MIME type. Could it happen, that cdn sends images with wrong MIME so Google interprets them as pages...?
    – Evgeniy
    Jun 6, 2017 at 15:40

1 Answer 1

Reset to default
1

Here are some things (not necessarily all things) to check:

  • Ensure all incoming http:// requests are 301 redirected to https://
  • Ensure all hrefs are updated, including hreflang, etc.
  • Ensure any plugins are updated with https
  • Ensure sitemap was updated properly
  • Ensure robots.txt was updated properly
  • Ensure htaccess file updated properly
  • In Google Search Console, add https versions
  • In Google Search Console, update parameter settings
  • In Google Search Console, update sitemap (only after confirming sitemap updates)
  • Ensure new CDN image serving does not include new session/version parameters, for example
  • Ensure new image serving has not affected alt tags and all image file names remain the same
  • Double check your canonical tags
  • Double check all old redirects
Improve this answer

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.