Virtual domains problem

Problems on two virtual domains which used to be mirrors of each other.

[bkuemmer]

Hello there,

we have two named virtual hosts on one IP-address which have been online for several years. Previously, quite a lot of content was available on both virtual hosts (e.g. all ~user directories), and Google started to think that one site (the newer of the two) is a mirror of the other.

This resulted in pages which only are available on the newer virtual host being indexed (and found in Google) with an URL pointing to the older virtual host.

We now started to disentangle the two hosts and have set up 301 permanent redirects for the content which was indexed with the "old" URL. It seems to be general consensus that this should convince Google to update the URL.

Our problem is that this does not seem to work. It seems that Google does not like the redirect old->new because it basically thinks that old=new (we could see this for pages which only had external links to the new URL but showed up in Google with the "old" URL which was never published anywhere), the result is now that redirected pages which were indexed are dropping out of the index.

Doing a site: search for the new domain in Google also returns hits for the "old" domain, this also suggests that Google internally considers the two virtual domains to be the same (this might be worsened by the fact that both domains resolve to the same IP-address) and doesn't even check if this is really the case.

Any hints on how to resolve this problem? Right now it seems that we will lose all the indexed content which was redirected to the "new" virtual host.

Thanks
Bernd

[claus]

These things take some time to get sorted out by Google. I'm sure you're doing the right long term thing using 301 redirects and collecting your content on one domain - both for your visitors and for the SE's.

I recently did the same, it's working now, although in my case it was abandoning a subdomain (www.) and not a domain. Recently means one month ago. Until it was solved, old urls' showed up side by side with the new, now it's only the new. I have not lost any pages doing this.

/claus

[bkuemmer]

Unfortunately, currently I am only seeing old URLs dropping out, but the "new" host is never crawled, I only see requests for / and robots.txt, not for any of the redirected pages.

To make matters worse, there is still content at the "old" domain, this is no complete switch of domains but more of a split.

The other strange thing is that when I search for "site:www.[my new domain].com -madeupword" I see the pages listed for my old domain. This really confuses me, why should a site:www.[my new domain].com search return results for another site?

[claus]

Well, somehow Google has got the impression that the "old" domain is the "right" domain and that the other is just a duplicate domain that also points to it.

It does not make it better that you have pages on the "old" domain that can not be reached through the "new". I'd rather wait untill you could make a "clean switch"; stop using the "old" domain completely and start using the "new" only, plus 301 redirecting old to new at the same time.

Or better (Google-wise) stop using the "new" and keep using the "old" as this one apparently has most links pointing to it. It' gets crawled deeper, so it must be seen as more valuable. As there is probably a good reason for the switch to another name, this is not an option, and then you should make the switch complete in stead. Consistency is key - you want to do the same thing with all pages, not something with some pages.

/claus

---

Btw: almost forgot that:

Welcome to WebmasterWorld bkuemmer :)

[bkuemmer]

> Welcome to WebmasterWorld bkuemmer :)

Thanks!

The problem is that we cannot abandon one of the domains, since they represent two different (but closely related) research institutes, the newer one is something like a spin-off from the older one...

But maybe Google is starting to catch up, last night I had the first Googlebot visit on the new domain on one of the redirected directories. Still, this page is in the index with the old URL... I think there's not a lot more to do than wait.

Thanks anyway for your suggestions!

Cheers
Bernd

[claus]

>> we cannot abandon one of the domains

Okay, i didn't know this. I just thought that it might be that you were in a company that had changed its name or something like that.

Then, if one institution is really a spin-off from another, i would draft a clear strategy for which part(s) of the total amount of pages would be located at which domain. That is, set up rules for when one domain, and not the other, should be used.

The aim would be to make sure you do not have duplicate content, and that the users will find some sense to the order and placement of documents (relating to the functions and history of the institutions perhaps). Search engines don't care if the order of your documents make sense, they just read them, but if the divisio between which domain does what is not clear, you'll end up confusing your users.

/claus