Forum Moderators: Robert Charlton & goodroi
AlexK, that's a different domain. But the point is very well taken. You've found a pretty obscure query (~295 results) that the keyword stuffing spammers like to target. I'll check this out in more detail.
[edited by: Brett_Tabke at 1:03 am (utc) on Nov. 12, 2005]
Is Jagger3 .7 or .9
I thought my question related to how the DC's worked, not how the Jagger update is affecting them. Ie, if a default browser is set to one DC, how entering that DC in the address bar can affect the results. I thought they were one and the same thing, if the default is X, then entering X in the address bar and searching should surely give the same results?
I'm really just trying to learn something, I'm not trying to be a pain.
I would keep your questions mono-silable if you want an answer from this crew. ;)
A good question and deserves a good answer, so who knows?
One way that you can discover which datacentre your "default" version of Google.com is currently using is to use traceroute. Open up your computer's Command Prompt (in Windows: Programs/Accessories) and type in
tracert www.google.comand read the bottom number after the trace is done.
[edited by: Patrick_Taylor at 9:37 pm (utc) on Nov. 7, 2005]
About 5 days. But although I believe it was the dissatisfied report that did it, I can't be totally sure.
[edited by: LegalAlien at 9:34 pm (utc) on Nov. 7, 2005]
When that happens again, we'll know what set of results they are using. Until then it's just guessing.
Is Jagger3 .7 or .9
Jagger is 9! The 9 results are temporarly now on 7!
That has not happened for about a year for me.
OK, For my main site perhaps something encouraging in the Allinanchors.
The Homepage is currently url only at the moment - despite Googlebot happily crawling internal pages on a daily basis.
However, on J3 Dcs 66.102.9.104 my homepage is ranked 56 for an allinanchor:company name search - on the 66.102.7.104 (J2 with a twist) I am ranked 35 on an allinanchor:company name search.
On DCs which are not J2 twist or J3 I am 828 on an allinanchor:company name search.
Now - when I first noticed this I assumed that it was the blending of 66.102.9.104 and 66.102.7.104 (as the allinanchor improvement happened there first) - but I take it Steveb what you saw on 66.102.7.104 does not look like it has made any impact on the 9.104s at all?
So might mean nothing.
When doing a tracert, the IP address is 66.102.9.147.
If you have downloaded the google task bar, and search from there, is the DC the same as it should be in the tracert? Or is there a default DC for this?
tracert www.google.caand see what datacentres are being used there too (but only for your own search results).
Me too.
>> What is the expectation for pages where both www and non-www are indexed with different cache dates and content, and also different PR, but where the 301 redirect is in place? <<
The expectation is that the www pages will continue to be indexed with full title and snippet, and that the non-www pages will drop out within a few weeks; but only for "normal" pages in the results.
For supplemental results, they are all "stuck" in a time warp. Unless Google cleans up the supplemental index, they will be listed forever.
>> Does anybody see any sign that the redirected pages are being - or will be - removed? <<
Nope. No removal. No change. See post #400 in this thread.
>> Still don't see any solutions for the canonical problem and the supplemental database is still the same <<
Yes. Nothing happened there at all.
Although serps will benefit from this, the pre-Jagger directory and toolbar PR updates will be totally out of sync. Three (or so) months seems a long time to leave it like this, so does anyone think we might be lucky enough to see an interim PR update?
[edited by: LegalAlien at 10:10 pm (utc) on Nov. 7, 2005]
is faster and gives the same result. Same thing, windows cmd window, type it in, hit enter.
btw. 'ping' and 'tracert' give different results.
thanks again.
joop
There is no "default google".
I know. The expression was used to characterise a particular user's current datacentre, even if, as theBear says, it means nothing.
[edited by: Patrick_Taylor at 10:06 pm (utc) on Nov. 7, 2005]
Even an IP address means nothing.
The server bound to an IP address can change as can the IP address attached to a fully qualified domain name.
[edited by: theBear at 10:05 pm (utc) on Nov. 7, 2005]
Whats your opinion on Jagger3 until now? especially when it ended (not as expected) less canonicals and less supplementals.
Thanks.
Not ended yet Reseller I hope. GG played down the Canonical fix a while ago though (not just today) :(
[edited by: Dayo_UK at 10:07 pm (utc) on Nov. 7, 2005]
And how about you friend. How do you see your sitīs chances now?
Thanks.
If I screw my eyes up really tightly I can see improvements - but it looks like Google have missed the mark on the whole.
This is my own feeling on Jagger3. I feel that GoogleGuy has expected more of Jagger3 than what he got now. Matt sound as such too.
This morning you can also read between the lines of GG the same theme.
But as I said, its my feeling, and I mightbe wrong of course.
"btw. 'ping' and 'tracert' give different results."
Yes, and if you ping 10 times a day google, you will get different ip addresses, same for tracert, both are simply getting the current ip of the domainname you requested.
ping is just faster because you don't actually have to do the whole tracert which in windows takes forever.
You'll notice that when you use tracert the first ip given is the ip of the domain you requested, same as with ping.
What you're seeing is your request get sent to different datacenters at different times.
Joop is going to totally confused if he listens to any of us.
Maybe a discussion about the backend db that a server actually looks at also not only changing but maybe coming from different db servers would be overload.
First they are there then they "ain't" or hey how did that other site get between us it isn't even on the .7. or .9. servers.
I'm pretty sure I understand now... thanks :-)
Only the wisest of the wise can discuss the deeper mysteries of db selection etc.
re Jagger though, it looks pretty decent. Although as gg or matt cutts said, this isn't one update, it's 3, that are somewhat interconnected. Calling them all Jagger will make it harder long term to really separate what each component did and does.
Future historians of the web however will look at the second 'update' thread, which may have 10 actual postings and 1000+ chatter and scratch their heads wondering why. But that's their problem.
It's somewhat unfortunate that more attention was not paid to each component, then to its connection to the the next, but that's life.
But results are pretty solid, some areas not great, I can definitely see where they are headed with this.
I wish them even greater future success in detecting and wiping out fake link networks.
>>>>
Joop, yes, you understand now. All the ip is showing you is what ip address the domain name you typed in is resolving to at the moment of that request, whether ping or tracert makes the request is not relevant.
[edited by: 2by4 at 10:39 pm (utc) on Nov. 7, 2005]
If you add -t (space dash t) to the end of your ping request, it will ping every second until you break it with ctrl c. I wondered if Google's IP would change dynamically during the session -- seems not! Stop and restart though and you get a new IP.
