How are you determining you have duplicate content indexing problems?
If it’s links for your own site then each linked to page will have the rel="canonical" meta data set to the non-query-param version so shouldn’t cause any duplicate content issues.
I had a quick look through a few of your posts and didn’t see any ?ref=linuxlock.org parameters on your links - are you able to share an example?
To completely turn off outbound link tagging (assuming that the Ghost core is to blame, not some javacript in code injection), go to /ghost > settings (the gear icon) and scroll down to the analytics section.
I’m not the OP, but I wonder if their concern is coming from the Google Search Console report. It’s normal to have some “Alternate page with proper canonical tag” entries in GSC, and these are the result of other sites linking to you and including a ?ref=other-site link. They’re basically harmless, and they’re out of your control, beyond asking the other site to not tag their links. (Yeah, good luck.) Those inbound links are still good links.
Ghost should not add a ref= link to internal (same-site) links. If you are seeing this, I’d suspect you’ve got something in your code injection that is doing it. (Be sure to check code injection header and footer.) Or maybe you’re accidentally linking to a domain name variation that isn’t the one Ghost is actually using. (If Ghost thinks your site is www.yourdomain.zzz but you are adding links to blog.yourdomain.zzz, that might lead to unexpected link tagging.)
Hello, I actually disabled it from ghost > settings > analytics section.
Actually I have installed ghost in domain.com/blog/ url so in case i link to a page from domain.com/something then it does not add rel="canonical" i confirmed that by inspecting the it had rel="noreferrer" instead.
Talking about issues, then I think it did, I am saw warning in Ahrefs web master and even in google webmaster All of my pages (including blog and other) were indexed, just the page which I linked too was crawaled but not indexed (google webmaster had no specific reason for it though)
Is it possible to turn off the ?ref parameter on a post-by-post basis? I have instances where the destination site chokes on any passed parameter it is not expecting.