Forum Moderators: Robert Charlton & goodroi
At one point I was trying to explain about the <base> tag and about 301 redirects and duplicate content. I thought that I would show some examples of sites that I know have "duplicate content" problems to clarify exactly what the problem actually is and how to recognise it in search results.
I did a site:domain.com search for a site that has about 400 pages indexed, and was told by Google that there were no results. I assumed I had typed the domain name wrong, tried again, and got no results again.
I tried the site:domain.com search for another site, one that also usually has several hundred pages indexed, and got zero results for that search too.
I tried a third site, yet another with at least a few hundred pages indexed, and it too reported zero pages now.
I tried another search that usually gives about 60 000 results and got "1 to 24 of about 24".
Seeing that something was seriously amiss I tried site:dmoz.org and got ZERO results for that too.
Just to clarify: the searches were done on www.google.com from the UK at around 21:00 UK time, through a major ISP.
At this point, the friend became bored, and started asking about some JavaScript code that hides email addresses from spammers and we went and looked at that instead... so I never did find out the IP address of the datacentre with no data...
Can anyone else find it?
Arriving back home I have tried these searches on several datacentres and have yet to see those results again. I use a different ISP, different OS and different browser, so may not ever be able to see them from here...
[edited by: g1smd at 1:09 am (utc) on Mar. 24, 2006]
If I key in:
www.dmoz.org I get a title, snippet and link to dmoz.org.
However, if I then click on:
Find webpages from the site www.dmoz.org or just key in a site: search there are no results shown.
If I just search on the string www.dmoz.org then the first 2 results are for dmoz (2nd indented) with a more results link which takes you to 29,900 results from DMOZ when followed.
Recently I've seen more and more of these and similar problems - one of my own sites (asked about this in a different thread) shows pages for the site command but shows no pages when using a general url search.
This is on a BD resultset and this DC has been switching betweem BD and non BD over the last few days.
My supposition is that G are rebuilding multiple indexes on their new software platform but are either missing a whole bunch of data or the data realtionships have somehow got skewed.
although the cached results in my searches show the above datacentre if I ping google I get
66.249.93.99 or
72.14.207.99
respectively for www and non www googles.
I tried site:dmoz.org and got ZERO results for that too.
If I remember correctly, dmoz.org used to redirect all the with-www urls to the no-www version -- maybe two years ago. I never checked out what kind of redirect it was, but I clearly remember watching the url change in my address bar, and thinking that they were gpoing in the opposite direction from what I would expect -- perhpas to keep the urls shorter by four characters. Then, for some reason, they began to resolve both versions.
If some data centers are using old data, I would think that site:www.dmoz.org would be the zero, and not what we saw a few hours ago. Curiouser and curiouser.
Google reads around 800'000 Pages from our sites per month - with little or no effect. Most of our content is not cached or has cachedates that are 10 months behind.
Traffic has gone down, Adsense earnings have been the lowest since 10 months and yes - we are getting angry here.
I do appreciate statements from MC regarding the update but I want to hear where our pages are. How in the world should we offer an adsense solution such as the "search on my page", if our pages are not available?
I like to hear a statement that says: "We keep the data we read - at least we keep it somewhere and it is not lost."
[72.14.203.104...]
On that same DC for site:www.dmoz.org
Your search - site:www.dmoz.org - did not match any documents.
Back in 2004, they both used to show tens of millions of URLs ALL of which were from other domains - a classic search to show the 302 redirect URL hijack in action if ever there was one.
Within 24 hours of mentioning those two searches in this forum, they suddenly showed zero results - blatently "fixed" by hand. You'll find that I posted about this several times back at that time, a time when 302 redirect URL hijacks were rife.
The search that played up last night, was the other one: site:dmoz.org and I guess there is only one datacentre showing that zero result for it. Every search that I did had way less results than expected at that time.
Not sure if it is related [prisonplanet.com] in any way.
