Help to find 404 links listed in WMT

[meelosh]

Hi All....i am stumped with this one.....in WMT i have a 404 that is linked from 10+ pages..all pages being my own pages on the site in question.
I have run a broken link checker on all the pages and all comes out clean.....i have fetched pages as google bot and do not find the link in question....could it be an error?
it is weird as the link is only a letter from root like http://www.example.com/a
I am hoping this is an error from google as i have done all i can to find this renagade link but come up empty everytime? any suggestion would be awesome..thanks

[tedster]

If you've done "Fetch as googlebot" for the reported pages, and they don't have the link on them - then it sounds like a bug. If you are using relative linking, then be sure that you are searching the source code for the right href values.

[meelosh]

thanks tedster...for the life of me i cannot find it..but google seems to think these pages have a broken link on them....hoping it is a bug.
has anyone else experienced something like this?

[meelosh]

This link is still being found on my sites pages on a daily basis....i have run all kinds of tests like xenu, fetch as google bot and do not find this link mydomain.com/a
I know @smace mentioned the same issue.....has anyone else seen this and does anyone have any recomendations on what to do...it is driving me crazy.
has not affected rankings or traffic but the 404 count is adding up by the day

[rowtc2]

Look for commented code too,to not be broken links there.

[meelosh]

the pages in question have javascript that assigns events to elements..
$("a[rel='widget']") could this be what the bot is picking up as a broken link? this is all i can think it is as there really is nothing else....i have been over these pages a hundred times

[aakk9999]

I have seen it on one of sites I am looking after, a 404 report in WMT for www.example.com/a (where "a" is actual page requested and not just an arbitrary example).

But this is the only "strange" URL requested.

You are saying that your 404 count is adding up by the day. Does this mean that you are having other non-existant pages reported in WMT 404 section apart from already mentioned mydomain.com/a ?

[meelosh]

sorry not 404 count....the amount of pages this 404 is being found on is increasing...i am wondering if it is the way gbot is trying to handle and probe the javascript?

[mercedesP]

Exactly the same here. WMT shows 404 crawl errors from several pages on my site pointing to "www.mysite.com/a".
In theory this was last updated on 21/11/2010.

I did all possible checkups to find that broken link, but it just doesn't exist and never has.
So, the only possible explanation is that there's a bug in G's end, as it's too much of a coincidence that the requested page is the same as meelosh's and aakk9999's.

[meelosh]

@mercedesP i am curious...did any of these pages have javascript that assign events to elements like this $("a[rel='widget']")this is so far the only difference i can find on the pages gbot is finding these links?

[mercedesP]

@meelosh

Nop, nothing! no javascritpts assigning events in my site at all

The only thing that I could "come out with" is that 2 of my pages begin with "a" ....

[meelosh]

Hi Guys...for what its worth..i think this is what google is picking up and reporting a 404 from your own pages...it seems to be a piece of code in jquery <a href="/a" style="color:red;float:left;opacity:.5;">a</a> that is adding the /a
I have found this on only pages with jquery and it is an older version.
The newest version changes "/a" to '/a' so i am going to make the changes and see if this helps.....even if it does not affect crawling or indexing or ranking i donot like "unknown" 404 showing up.
will let you know if this works