What might make Bing.bot find a URL that looks like this on our site?

[danatanseo]

I have been doing something Richard Baxter recently suggested and reviewing our server logs.

I have found an oddity that hopefully some of you smart Mozzers can help me figure out.

Here is the line from the server log (there are many more like this):

157.55.32.166 - - [04/Mar/2013:08:00:59 -0800] "GET /StoreFront/category/www.ccisolutions.com/StoreFront/category/shure-se-earphones HTTP/1.1" 200 94133 "-" "Mozilla/5.0 (compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm)" "-"

See how the www.ccisolutions.com appears after /StoreFront/category/ ? We used to see weird URLs reported in GWT that looked like this, but ever since we fixed our canonical tags to be absolute instead of relative URLs, they no longer appeared in our Webmaster Tools reports.

However, it seems there is still a problem. Where/how could Bingbot be seeing URLs configured this way? Could it be a server issue, or is it most likely a data problem?

Thanks in advance!

Dana

P.S. Could this be resulting from our massive use of relative URLs all over the site?

[StreamlineMetrics]

I'm willing to bet that on some page of your site, there is a link pointing to www.ccisolutions.com/StoreFront/category/shure-se-earphones which is missing the "http://" at the beginning. So if Bing or a user tried to click on that link, they would be directed to /StoreFront/category/www.ccisolutions.com/StoreFront/category/shure-se-earphones instead of the correct link.

[danatanseo]

While this is certainly a possibility, I'm not sure it's the cause of the problem. If this were the case, wouldn't it most likely cause a 404 error, instead of rendering the proper page (albeit with a very funky URL) and a 200 status code?

The other thing making me think it's not just a poorly constructed link on the site is that there are over 100 of these in the server log, from just one day.

Thoughts?

[StreamlineMetrics]

Can you tell from which page Bing is trying to access these URLs? And it only happened on the 4th and not on any other day? Could it be an issue with the sitemap on that day?

I'm looking at your site now and the page http://www.ccisolutions.com/StoreFront/category/www.ccisolutions.com/StoreFront/category/shure-se-earphones is returning a 200 response code to me, not a 404 code. The key is to figure out how Bing discovered the URL in the first place...

[KeriMorgret]

Dana, you might also want to contact Bing at https://support.discoverbing.com/eform.aspx?productKey=bingwebmaster&ct=eformts&scrx=1\. I sent a quick note on Twitter to Duane Forrester and that's the URL he provided.

[djlaidler]

Hi Dana,

I agree with Streamline, there will be a hidden issue in you site that it attempting to connect to an under formed link (a URL missing 'http://'). Given there is a number of them in one day I will guess this is happening in a templated page.

Have a look at;

http://www.ccisolutions.com/StoreFront/category/www.ccisolutions.com/StoreFront/category/shure-se-earphones

It renders as a page.

The best course of action would be resolve it at the source. If you can pinpoint when this issue is due to occur next, have your developer get each page to append it's URL into the log at the beginning of the page. Then you should be able to determine where the issue is occurring. I am hoping you well see a discernible pattern.

Worse case scenario, possibly a canonical will work, OR create a REGEX redirect to handle this URL pattern in htaccess...

Hope this helps,

Dan

[danatanseo]

Thanks Keri. That's very helpful. I will do that.

[danatanseo]

Thanks to you both Daniel and Streamline.

I believe the problem may have to do with our .htaccess file. I am obtaining a copy of it now.

[danatanseo]

Hi Streamline,

I thought I would circle back and update everyone as to what I found. You were correct about mal-formed URLs being the culprit of this problem. We have many isolated incidences of URLs for internal links that are missing the "/" at the beginning of a relative URL. There are inconsistencies on the relative URLs all over the site. It's certainly an example of one of many problems that can be caused by using relative rather than absolute URLs.

Since we are in the process of completely re-doing the site and moving to a new platform, it's something we can definitely work to get right during the transition.

Thanks again to you, Daniel and Keri for jumping in with answers.

Dana