For our site we set up a CDN url using a CNAME pointed to our main domain. Unbeknownst to us, this caused a very large pool of duplicate content, since google could crawl the CDN and reach all the URLs that should only be reachable through our domain.
This has also caused nearly 90k backlinks to our main domain, as all of our pages link to the domain in some way. Obviously, this is unintentional but we fear it looks bad to Google.
We attempted to solve the issue and the CDN now returns a basic HTML format with
<META NAME="ROBOTS" CONTENT="NOINDEX, NOFOLLOW">
<head/> for non static files.
We'd hoped this would lead to Google de-indexing all the pages and the 90k erroneous backlinks would disappear. We attempted to use the disavow tool for these purposes.
We've seen the number of indexed links from the CDN drop off (according to WMT, it's actually at 0, though some do return in a search), but all of the backlinks remain.
We're not sure what to do as the issue was unintentional and we'd ideally like to find a way to tell Google we do not want any credit from those CDN backlinks, but we have no better ideas than making the CDN itself noindex.
Any ideas would be appreciated, thank you.