Cache version of page shows a different root domain

[latimer]

We manage content for several sites that have same folder and file names.

finding some of the pages in google index for one of the sites by doing a search for:

site:www.sitenumber1.com/folder1

example of page found in the search results:

www.sitenumber1.com/folder1/file1.cfm

clicking on the cache link for the above example page and the address showing in the address bar is:

www.sitenumber1.com/folder1/file1.cfm

but when the cache version opens it shows at top of page in the grey cache info area a different root domain with same folder/file path as:

www.differentsitename2.com/folder1/file1.cfm

and the cache shows the www.differentsitename2.com version of the page. As mentioned the address bar refers to the www.sitenumber1.com page.

clicking on the "current page" link at top of cache page brings up the www.differentsitename2.com live version of page.

we expect that sometimes google filters pages out allowing the site with the most quality score or some other measure to have the page included, but showing cache versions of pages with completely different domains and visual content doesn't make any sense.

We did add a g-sitemap to one of the sites some time ago and wonder if somehow that causes something like this to happen, as it has never occurred before and these sites have been out there for 5+ years.

Has anyone else seen anything like this, or have thoughts on what is causing this? We removed the g-sitemap from the one site, think this will help?

[tedster]

I've seen this kind of cache confusion only a few times - and each time it involved a parallel situation ('umbrella' hosting for the crossed-up domains, similar templates, and so on). I've never pinned down the exact reasons - for instance, is it shared IP address that contributes to the confusion, or are there errors in a generated xml sitemap, or even server response problems when googlebot makes requests.

I'd say check each of these areas first, to ensure that there are no technical errors you can fix. Then move on to interacting with Google about it, first through Webmaster Tools accounts and if that doesn't sort it out, possibly take it to Google Groups.

[latimer]

thnx Tedster. Gives us something to go on. Any other comments appreciated.

[jimbeetle]

I've also seen similar in search results using the site: operator. In these cases it usually involves either new sites on shared IPs, or a new site on a dedicated IP that had recently been used by a different site. In these cases Google usually gets things cleared up in a week or so.

[caribguy]

I have seen 404 errors for things like:

www.example.com/page1.ivnu
www.example.com/page2.ivnu

Which turned out to be the result of Google queries for 'example'

I later found out that a non-related site www.example.biz utilizes a CMS that creates pages with the ivnu extension.

Used WMT to remove those non existing pages from its index.

[latimer]

The sites in question do share ip address. We are considering giving them seperate ip addresses. the xml sitemap was correct, although we are removing it anyway to play it safe. Fairly certain that our server is not the issue. Any other suggestions?

[waynne]

Do the description tags match on these pages?

[gusp]

latimer, I had a cache confusion situation last year.

I had two domains that I had parked on my main domain with 301 redirects.

I removed the 301 redirects and put up sites on the domains.

But Google's toolbar cached snapshot views for each of these two domain's index pages persisted in showing a view of my main domain's index page for months.

The HTTP headers of the two domains no longer showed 301 redirects and there was nothing wrong on the host end or at the domain registrar.

Their content was totally unique and their templates were all different.

I tried using the no archive Meta tag along with changing the content of the index pages then waiting until the sites were crawled and the toolbar cached snapshots were gone.

Once that happened I removed the no archives Meta tags and waited for the index pages to be re-crawled.

As soon as that occurred once again Google's toolbar cached snapshot views of each of these two domain's index pages reflected that of my main domain's index page.

It was driving me nuts. It looked like I was cloaking.

But here's the catch, in my situation when I used the site: operator the cached link views for these two domains correctly reflected their own content whereas Google's toolbar cached snapshot views did not.

Three months later when Google did a toolbar update the entire situation cleared up.

[szykman]

Hi latimer

if you have many sites in the same folder (what in cpanel servers are called addon domain) or if you have domain alias or also called parked domain and your server had a problem at it's configuration this could be caused your issue...

also, is there any chance that you have a [example.com...] in a global shared ip adress and then every site from your server is pointing to http://www.yourmainsite.com ?

well it have being happening for me since Sep/2008

all Page Rank from my Keywords and SERPs were "stolen" by a weak web site...

ir is called deceptive redirect usually made by a 302 redirect

now cache:https://www.weakdomain.com/page shows www.mysite.com ! what is awful, really bad

i change the ips and placed 404 pages for all pages coming from the other site and almost 2 months has passed and google still showing a couple of my authoritive pages in another sites, consequently my keywords are not performing well on SERPS

if you are in a windows server maybe a misconfigured Virtual directory, or in any server a misconfigured subdomain or alias ?

try also to check searching like this:

www.siteA.com "keyword"
www.siteB.com "samekeyword"

this can give you more clues

regards
Claudio

[edited by: tedster at 6:01 pm (utc) on Nov. 22, 2008]
[edit reason] de-link the example url [/edit]

[latimer]

an update to this ongoing situation:

many of the cached cross over versions initially found back in November are still in the index. some have fallen out.

the problem at that time was occurring on 4 or 5 sites all on the same ip. We dedicated an ip to one of these sites and blocked google from the rest.

We also have a number of others sites with same content and folder path names on seperate ip.

We had not seen this problem occurring across ips until after the first of the year. We have now found at least one example where the cache is crossing over between domains on seperate ips. This occurrence has cache date of 1-5-09.

One other observation, Webmaster Tools is showing pages that exist on several of the crossed over cache sites as external links into the one site that we didn't block google from that is set up in WMT.

will this cache cross over cause page rank cross over among the domains? Any other thoughts on this? Anyone else seeing this type of thing?

[tedster]

Have you considered adding the <base href= > element to the head section of each page? It might help sort things out if you place each page's absolute url that way.