Weird 404 Errors

[G2W]

Hi All,

Although my Moz error scans have been pretty clean for a while, a law firm site I manage recently cropped up with 80+ 404 errors since the last scan.

I'm a little baffled as the url it shows being returned looks like this:

http://www.yoursite.com/ http://www.yoursite.com/resource.html

For some reason it seems to be initiating a query to call the root domain twice before the actual resource.

I installed ModX Revolution 2.2.6-PL on the site in question, and am hoping a canonical plugin I just started using will take care of these.

Has this happened to anyone else? What did you do to solve the issue?

Thanks for your time and any tips!

[brad.s.knutson]

%0A is the line feed character, so it looks like your CMS may be spitting out links that browsers and crawls interpret as relative links.

If your link appears like this:

[The link will be interpreted as relative and result in the link that you found on your Moz error report.

It's probably a problem with how the CMS is spitting out the href attribute, but it's hard to say without knowing more information.](%0Ahttp://www.yoursite.com/resource.html)

[seoagnostic]

Bradley's response is spot on. I coincidentally manage a large site in the legal area that has had errors like yours although the error isn't law related! As Bradley implies, this is typically the unintended result of code that is spitting out the unintended line feed in the CMS. I'm guessing that what you're seeing is probably the result of something related to PHP rather than an error in user input. Typically entering white space into user editable areas will result in it being stripped. When you have actual code like this inserted, it's the result of some line of PHP someone edited and saved without realizing the effect. I've had this happen before with RSS feeds where one little glitch will put a forward slash to the end of a URL and connect the beginning of another. Good luck with finding the solution, which shouldn't bee too tough.

[danatanseo]

Yes, I have seen this problem before. Bradley and Michael are both correct in that it had something to do with relative versus absolute URLs. In our case, it was being caused because we had relative URLs in all of our canonical tags. As soon as we fixed them to absolute URLs the strange looking 404 errors went away. Hope that helps!

[Christy-Correll]

Hi David, were you able to resolve this issue?

[G2W]

Hey Christy,

Nope, Never did figure out an answer!

I took a break for a while but now I'm back trying to divine a solution.

Re: Dana's suggestion, I did make sure that our canonical plugin was using absolute urls, but it looks as though that did not solve the issue.

Not sure how to locate a potential PHP glitch that might be causing this... any pointers?

[Christy-Correll]

Hi David, I apologize for the delayed reply. I'm going to check with some other Associates and see if they can help trouble-shoot. In the meantime, please let us know if you have any updates. Thanks! (Christy)

[evolvingSEO]

Hi David

I would recommend cross checking this with Screaming Frog and/or Webmaster Tools. It's only a concern really if the web spiders and users are experiencing these 404's. I have seen it happen where Moz's crawler may hit 404's that Google and/or users do not.

If you get the errors in Screaming Frog or Webmaster Tools as well - here's what you need to do to fix them.

1. Go to the source page of one of the broken links.
2. View the HTML source.
3. Do a control-F and search for the broken link (have it copied to your clipboard)
4. Determine where in the code it's coming from.
5. Then you can probably debug it from there.

Let us know if that gets you there.

Thanks!

-Dan

[G2W]

Hey Dan,

I had kind of assumed that it might be a false alarm from the Moz scan. I typically use Xenu to check for broken links periodically and it hasn't shown any.

Thanks for the tip!