Forum Moderators: open
I also notice it seems that Google has 2 banks of datacenters. Only one of the 2 does the partial update. Next partial update, the other bank of datacenters is used. Looks to me like -ex, -in, and -zu are involved this time. And, possibly -va. However, it could be that all of these are being rerouted all to one physical datacenter. Looking at traceroutes this may be the case.
Do you guys have also notice the slight difference in adwords result between data centers?
Estimated Google index
ex - 1,950,000,000
in - 2,250,000,000
cw - 2,240,000,000
dc - 2,240,000,000
fi - 2,230,000,000
va - 2,230,000,000
sj - 2,240,000,000
ab - 2,240,000,000
zu - 2,240,000,000
One thing I have noticed is that the people in this thread have been talking a lot about the changes in the DC's yet I've seen very few mentions of the GoogleBots.
I would imagine that some observations of bot behavior tied in with tangible results in the Serps would support the theory proposed by rfgdxm1, rather than watching just the dc's (which we all know by GG's comments are under flux for the period of the Esmerelda update and some time shortly thereafter).
WRT the bots, while there are some fresh-dates out there, there seems to be - in my neighborhood - less of them. To support this I have seen less GBot traffic on our sites and our main site in particular.
The observation wrt our main site may be a bit skewed, as a redesign went up this weekend, and it could be that GBot doesn't like it, as all other bot traffic is up, but Gbot is down, but it could also be that they are trying to balance the polishing of Esmerelda with getting new data into the results.
My 2cents, fwiw!
In May, it crawled and updated for all over the month and similar thing is happening once again in June. Also, according to googleguy, freshbot is behaving like deepbot. So, many sites are been crawled and updated every 2-3 days.
As far as I have noticed, this depends on the links to the website from the relevant sites and not the page rank of the site. If you get links from websites that are top rated by google such as different directories, universities, microsoft, etc., or yours is a news site, your website have more chances to be crawled and update on a monthly basis.
I have a website that has only 3 links from universities, dmoz.org, google directory and is being updated every week.
Similary another that has many links but not all relevant to that business, is updated 1-2 a week.
So, I agree with the statement "Google now just doing a continuous, rolling update."
next anomoly noted is that the fresh cache copies of some index pages are gone and replaced with old (month ago) pages
Yep, I see this too. We haven't been crawled lately, but there was a relatively-fresh version of our index page up before the recent round of f-tags (jun-24).
Anyway, I wouldn't call this an anomaly, I've seen this behavior on Google long before any of the recent changes started taking place.
<snip>I saw the light</snip>
You guys need to get out a little more. As Lou Reed said 'take a walk on the wild side...'
I'm not an "adult webmaster" - far far from it. But while you guys focus on 'blue widgets;' and 'neat widgets' - you potentially miss the real issues. The real action is on the darker side of what people want to sell.
Try canadian betting in www-fi
234,00 results - I'd say that's reasonably competitive - so read the 'keywords in context snippets'.
Try disney cartoons in www-fi and look at number 10 - the nude ones.
Try a search on long tongue on www-fi or www-in
Hmmm.
How spammy is that? Hasta la vista late 1990's.? Top 10 results are absolutely pathetic - any searcher should be appalled. Search for betting - you get pop up porno. Look at the url for result 3 for long tongue - great URL! And I've now locked my PC - if google isn't dafe for searches for Disney cartoons - then the internet - or Google - have lost the plot. Lost it.
I said yesterday that -fi apparently had no spam filtering applied - whereas www-in did. I take that back.
Clearly - its all crap. All of it. Whats an adult filter?
Hasta la vista google?
yahoo and google are switching data between the first group of datacenters (fi, dc, va, ab) and second group of datacenters ex, cw, in for last two days or so. aol is just showing data from first group. sj and zu are redirected to other data centers.
The most recent and best updated centers, IMO, are "in, ex, & cw"
This may be a stupid question, but are these datacenters geo sensitive?
If the information here [webmasterworld.com] is correct, then the data centers 'in, ex, & cw' are in California (sj is also but redirected) the home state of Google. The data centers 'dc, va, & ab' are outside California (and so is zu, but that one is also redirected) and fi is somewhere in America. So it could very well be geo sensitive.
Its not up to google to censor the internet (thankfully). They do censor sites from their results in some cases, but I don't agree with it in any case. Just taking on the charge assumes some responsibility for what's displayed for a given query, and exposes google to some level of risk.
Actually the site in question is quite spammy as far as doorway pages and that sort of thing, so if that's what you were referring to, I agree, the spam filters should catch that.
For those who are unfamiliar, there is a significant amount of porn out there that involves Disney characters. These porn sites coming up on a search for Disney cartoons are indeed relevant.
spammy as far as doorway pages and that sort of thing, so if that's what you were referring to, I agree, the spam filters should catch that
Yep - thats it. Just forget Disney - sorry I mislead you - of course thats a relevent result. As I said - hundreds of serps - sometimes I just get confused.
So forget Disney - just look at canadian betting - and try the keywords in context and the cache! Sometimes - you just can't cache good spam!
Absolutely relavent results. Nah - Googles not broken. Its better than ever.
: )
It seemed as though the index was shaping up on fi/cw/in. Now it is reverting back to dominic?
We have been calling it "One to None"- referring to pages that go from #1 on many of the datacenters to off the map the next day.
If this does settle down like the "traditional update" GoogleGuy said we were going to have, I liken it to musical chairs.
When the music stops you better hope it is on an hour where your site is listed.
I thus think the only "new" thing happening is that we're seeing more off-page factors considered fresh; and one bot name
As of the moment I am writing this, I still see the good results on those datacenters. I wouldn't be surprised if this doesn't last, though.
>If this does settle down like the "traditional update" GoogleGuy said we were going to have, I liken it to musical chairs.
I'm wondering if this will settle, or things will just keep eratically switching? At the moment, if you don't like the Google SERPs, wait a few hours and see what they are like then.
A new page, put up on the 18th was crawled about the 21st, appeared in yesterdays SERPs on all dc's, with no freshtag, then today is gone on all dc's.
It was #1 for the <title> kw1 kw2, so very cool, but it was a very brief stay in the serps for a freshbot hit... I'd usually expect 2-3 days.
I haven't seen 64.68 at all the last 2 days. I'd like to see this update truly over then find out what kind of stuff deepfreshbot has.
It seemed as though the index was shaping up on fi/cw/in. Now it is reverting back to dominic?
Yep...fi is back to Demonic for my SERPs.
1. There is now only one bot: it is crawling fresh & deep: it does certain pages everyday, but it does very deep pages only on certain days
2. Pages already in the index are being updated every contnuously and a lot more "off page" factors are being considered. I have been seeing pages that are in the index moving up and down a bit, but no new deep-page has been added. Pages I add with links from the homepage are added immediately, but deep pages are not being added...this is the same a before.
I can't get new pages spidered past the homepage for a brand new site of mine. I keep getting new links to it, but still none of that deep action. I'm just about down to actually submitting those pages...
newish site, 3 months old. index page spans to 10 or so category pages which span to 150 or so product pages. The product pages were crawled and indexed about 2.5 months ago but were only about 10 then, since they have dropped out and all of the new ones have not been indexed and the googlebot is just not going that deep into my site to index them. I even tried adding links to a few of these on the homepage but it will not follow links to my product pages.
A deep crawl, a deep crawl my kingdom for a deep crawl.
When will this madness end, this has to be the worst time to launch a site in the last 3 years.
Assuming this is the "traditional" update that GG alluded to, this period is taking so long that the temporary/flux results are becoming quite important. 10 days so far and no end in sight.
If updates take half the month, the dance becomes as relevant as the final results. The saying "Wait till the dust settles" is outdated.
Also, Google has been known to apply spam filters towards the end of an update. If this is still the case, one may be able to gather quite a lot of traffic before the hammer falls now.