Forum Moderators: Robert Charlton & goodroi
For those of you who may be seeing strange results on the following DCs:
72.14.207.99
72.14.207.104
Do a site:mydomain.com check for your site & check the pages that are returned.
Do the serp titles & descriptions returned look close to being identical (eg. site-wide title and description tags maybe)?
I have an idea brewing :-)
[edited by: tedster at 7:39 pm (utc) on April 17, 2006]
[edit reason] start new thread [/edit]
Also I'm in the same boat of losing pages from the index in the last couple of days.
"By and large, i believe that's what they're doing, yes."
With all due respect, your comment neither reflect knowledge nor respect for other members.
[edited by: reseller at 9:03 pm (utc) on April 19, 2006]
That is a generic error message that often is not a true reflection of what is happening.
.
Real example:
A page had an incorrect telephone number on it. The error on that page was corrected more than two years ago.
If you search for the new number you get a snippet with the new number and the cache is 10 days old and shows the new number. The result is normal.
If you search for the old number, you get a result that is Supplemental, and the snippet shows the OLD number. The cache is still just 10 days old, however the cache shows the message that you mention.
In both cases the returned result is for the exact same URL. There are NO links with the incorrect telephone number (unless you count the click from the Google SERP).
The error message isn't telling the whole truth. It should say: "these words were only in an older version of this page; they are nowhere to be found in this more modern cached copy of that same page".
[edited by: g1smd at 9:33 pm (utc) on April 19, 2006]
<Fixed typo in second IP address>
[edited by: tedster at 6:28 am (utc) on April 20, 2006]
Wow, im so excited i just cant contain myself....
Amazing, after the number of years that Google have been involved in search they still want to tinker about and "try something else"
What would be real nice and an absolute first would have been if a Google engineer had said. " We have decided to concentrate our efforts 100% on the existing search, fix cannonical issues, bot problems, 301s, 302s, indexing issues, PR issues and get the index working properly before we tinker with all your business interests and try soemthing else"
Sorry, but im glad i wasnt at pubcon i might have wanted to tip my beer on his head!.
Meanwhile if mcdar is showing a roll out of data from that test centre then the google serps are about to get even worse than they are currently which is hard to belive.
The serps in that data centre are just all over the place, i would go as far as to say that gigablast is delivering better quality serps than that data centre, its just a total mix, its one thing trying to be be smart and doing something new but they are making a dam mess of the entire index, gawd knows what they are tinking with delivering those results just looks like, chuck everything in and mix it round!
What a waste.
Good morning Folks
Another new shining day starting at a Cappucinno. Life is both wonderful and great :-)
Well I see some chnage on [72.14.207.104...] which might look promissing. The top 10 sites for my test search query ALL are relevant. And Its affiliates friendly too.
Wish you all a nice day and happy Google DCs Watching.
>> Well I see some chnage on [72.14.207.104...] which might look promissing. The top 10 sites for my test search query ALL are relevant. And Its affiliates friendly too. <<
I too agree with you. You checked for relevancy and affiliate-friendly. And I checked for Canonical Issue.
The DC 72.14.207.104 solved the problem of Canonical issue. I say this because, my site used to have 9 results for "site:domain.com -site:www.domain.com". Now it shows no results at all.
By this one can say "72.14.207.104 might look promising" ;-)
In any case I see no progress on the canonical front. Every day's update has more or less 301s or duplicate pages listed. That doesn't mean much of anything.
Canonical issues are "fixed" when no inaccessible urls are indexed, when non-existent pages are removed from the index, and when pages are ranked properly. Google periodically hides some things, which of course is no solution to anything.
==
Just got a different set of data on 72 (not better or worse, just different), so perhaps it is in significant flux right this minute.
64.233.187.104
I hope those are not the shape of things to come. If so, looks like it's time to break out the check book and buy a boat load of links because those who have are ranking very nicely on that DC.
Hi,
For me I would say exactly the opposite. 64.233.187.104 is less about in bound links than the majority of current DCs SERPS in our niche. Looking at the first two results pages (prefs set to 10 per page) and I see most of the sites of the major companies in our field listed there with very few that I would not expect to see.
There is one page that I did not expect to see and that is on a secondary site of ours. It has just one link to it when I search <link:url> but it is relatively highly optimised on page, semantically rich and only has links out to relevant sites. This site has not been in the 1000 for our main target search term since Florida but has now appeared at #20 for that term on 64.233.187.104.
FWIW IMHO I would therefore say that 64.233.187.104 is a subtle shift away from off page factors (such as inbound links) and more towards on page factors such as relevance, semantic richness and outbound link relevance (link text using that term to an offsite page on that topic).
Sid
Anyone tried a site: search on 72.14.207.104?
It seems to still be showing a fraction of actual pages!
Absolutely - if you're talking about pages displayed - on us it's only 127 pages - 2 weeks ago we went to 258k.
And per my previous email, I'm seeing some of the DC's returning huge supplementals listings again -but i see no one else if reporting this - strange because it looks widespread.
As of today, it is returning only 900 results, but they are still not "exact match" returns and there are very many, very old, supplemental results in the list.
For some searches the old supplementals (from Jan 2005 etc) are gone; but for others some newer supplementals (cache January 2006 (and some from April 2006!)) have appeared.
As before the results are very wierd.