Why moz pro detects inexistent links?

[David-Kley]

Is there a reason you leave index.php in your URL? Might make it easier to strip it off using HTACCESS so you can see more clearly what you are dealing with.

[Er_Maqui]

There's no reason to leave index.php in different URIs, simply i have defined all links with their name, for example, for a root page from a section, the uri are defined as /project/index.php, not /project/ only.

I can clear them, or leave it. There isn't the problem. My problem are the moz stats returning non-linked inexistent addresses, and i doesn't know why.

Also, i've detected a failure who makes moz to find duplicated pages (A erroneous link with only a parameter) and i've corrected already. But i cannot find the reason for the inexistent pages.

[KeriMorgret]

Have you downloaded the CSV of your crawl report? You can look at the column for the referring URL and see what page is pointing to the 404 error.

[Er_Maqui]

I've requested the CSV and downloaded it. But i cannot see the page pointing to the error, i can only view the error:

That's the report:

<colgroup><col width="392"> <col width="28"></colgroup>
| http://maqui.darkbolt.net/project/chat/index.php | 404 |
| http://maqui.darkbolt.net/project/docs/index.php | 404 |
| http://maqui.darkbolt.net/project/down/index.php | 404 |
| http://maqui.darkbolt.net/project/foto/index.php | 404 |
| http://maqui.darkbolt.net/project/news/index.php?news=1 | 404 |
| http://maqui.darkbolt.net/project/project/index.php | 404 |
| http://maqui.darkbolt.net/project/ro/index.php | 404 |
| http://maqui.darkbolt.net/project/who/index.php | 404 |

Obviously there's all erroneous, the section "who" isn't inside "project" one. All links are valid without the part of /project.

I cannot understand why system are reading these links, on page, the links works ok.

[KeriMorgret]

There is a column to the far right that should show the referring URL. Be sure to scroll until you find that, and then you will see where we found those URLs.

Clarification: this is in the crawl test report. I'm not sure why you're seeing 404s in the link analysis page.

[Er_Maqui]

In the crawl test there isn't have the 404 errors.

[KeriMorgret]

So the link analysis is showing you that there are sites linking to pages that don't exist on your site?

[Er_Maqui]

Yes. Link analysis shows pages as: comusys.php/comusys.php (Who, obviously, doesn't exist).

On crawl analysis CSV i cannot get these pages. You can also get the analysis from your own open site explorer and view these pages doesn't appear.

And, now, Link analysis doesn't show any 404.

I've attached some examples of my campaign. I cannot understand why they are detecting this.

Any help will be apreciated.

Thanks,

Screenshot%20455.png Screenshot%20456.png Screenshot%20457.png

[SamWeber]

Hi there,

We're a bit lost in what you're trying to ask here. Based on what I've read, it sounds like you're saying that the weekly campaign crawl (not the Link Analysis data) is reporting 404's and you're not sure how/why that is happening. Is that correct?

[Er_Maqui]

Yes, that's correct.

[SamWeber]

Gotcha.

We're not actually reporting 404's in this case. We're reporting that one page is a duplicate of another which happens if the content on the source code is 95% similar or greater. The pages we're reporting that are duplicate did exist at the time of the crawl which is why they're showing up. If you made any changes after the crawl, there is a chance that the pages no longer exists in which case the next crawl will not show them as duplicates. They will be reported as 404's though so you'll still want to resolve that problem.

Outside of that, you can download the crawl diagnostic csv to get a list of referrer URL's. This is handy if you're ever unsure of how we got to a specific page. Hope this helps clear things up!

[Er_Maqui]

As i've told at first post and anothers,

1º: I have a failure of duplicate content, detected on first crawl. The dup content comes from a parameter on all pages (?login). I've solvented it making this parameter a link to main page. This change resolves the duplicate content on all site.

2º: On first crawl the system itself detects 404's. The do not exist. I haven't change anything to resolve 404's. If you see the urls, from my screenshots, they are strange URLs, because they detect the 404 on urls as style of index.php/project.php. This page doesn't have mod_rewrite or anything similar, because this, these URL's are impossible.

3º: I've tried to download the crawl diagnostic. On one file i doesn't have the referrer URL, on another, they doesn't have the 404 ones.

I'm trying to know why the system detects this pages when they doesn't exist and aren't linked from anyother site. If i have something bad, then, i have something VERY bad and i need to resolve right now. If not, i think the system detects some incorrect at my page but i cannot understand why.

That's worries me a lot, because, this page and campaign are a test. It's my personal web and i have only a few pages and a few links. But if i can't understand the results from this page. How can i understand / read the results from my best page, who haves more than 10.000 pages, multiple domains, social medias, and more?

[SamWeber]

Hi there,

• I reviewed several pages that we're reporting as duplicates and none of them are canonicalized. If you don't want to change the content in the source code so it's not 95% similar to the other pages, you'll need to add canonicals. The Help Hub has some good info on how to do this. You can also run a search in the community.

• If we reported 404's in the initial crawl, it's because they existed at the time. The most recent crawl isn't showing any 404's so this shouldn't be an issue anymore.

• Again, there are no 404's reported in this week's crawl for your campaign so there will be no 404's in the crawl diagnostic csv. That is where you'll want to go if this comes up again though.

-If the page on your site is linked to from anywhere on your site, we will crawl and report on it up to the page crawl limit set for the campaign. We're not going to report data for non-existent links as that isn't physically possible. I hope this helps clear things up.