Update Dominic - Part 8

Speculation Continues

[lasko]

Google now shows on their main index the new update and new algo.

I thought Google where going to add more back links but they no.

The number of backlinks on one of my sites have dropped by 50%

I hope this posting is acceptable

[finditnow]

New results in California.

[djgreg]

drewls: that could be true
carlr: for me there are eminent differences between -sj and -ex for example. If you would have lost all your top positions in an index which is said to be the new main index you would panic as me , wouldn't you?

[merlin30]

And just to add to the DNS theme - when your browser uses DNS to resolve the URL it probably caches the result so once you've hit say -fi or -sj you'll probably see these results consistently for several minutes until the Time To Live expires and your browser does a new query and (probably) gets a new address (and yet another instance of www).

Furthermore, sites like google-dance highlight about 12 data centres (including www2 & www3) - I've been doing individual pings today on www.google.com and found about half a dozen different IP addresses and none of them are those listed in google-dance. The question is just how many data centres are there? Perhaps hundreds.

[NicoleA]

One of 3 new sites I put up during the last crawl has been assigned PR as of this morning. Disappointing PR, but it's definitely there. Yet, it still doesn't show any backlinks on any of the servers.

Just thought this was noteworthy.

[carlr]

djgreg, I know there is a differennce between -sj and the different other servers - I just meant that I havent't seen anything happen on www from here, and from the very beginning of this "update". I have only around 300 visits/day on one of the sites i monitor so it is easy to spot differences.

[Critter]

This is the wackiest "non-update" I've ever seen...

This month's poll should be "In what thread will the real update be entered?"

My bet is on Update Dominic - Part 276.

Peter

[lasko]

I agree but in each thread you will find their is something unique thats happening that we will never be able to understand

This may not be a update but the fact is that this is the first time that some have seen results on the main index.

[ciml]

So, the www-sj and www-fi datacentres continue to be the only ones with the "different" index.

For the moment, www2 and www3 mostly carry those results, and www does sometimes.

Fair assessment?

[cobraq]

Seems like we will have to wait a few days more before the dance it finished... The Google update tool is currently doing some futuristic checking a couple of days into the future:

Last Check -177128 secs ago.
11:16 pm on May 14, 2003 (cet +1) :)

Hmm, the clock for posts has hang too. Anyone noticed that the dates and times for posted messages are jumping between 12 and 14 May too...?

Last Check -176461 secs ago.
11:16 pm on May 14, 2003 (cet +1)

[TheComte]

The cache seems to be missing on 2, 3, and sj. Is this new?

[wackmaster]

<So, the www-sj and www-fi datacentres continue to be the only ones with the "different" index. For the moment, www2 and www3 mostly carry those results, and www does sometimes. Fair assessment?>

From our seat, yes, except that we mainly see w2 and w3 continuing to point to -sj, not -fi...

[ciml]

Just date problems [webmasterworld.com], cobraq.

Incidentally, while the WebmasterWorld clock was set for May 14, I noticed that only www-sj and www-fi had the new index. Therefore, I suggest that we all just take a day off and head for the beach or something...

[Jon_King]

still only on sj and fi for me

[lazyz]

Yes - I will finally buy off on the update... but I am finding my old domains that I know, no longer exist and haven't existed for two months still in the data... what gives?

[Powdork]

Didn't GoogleGuy say he wasn't sure if they would move the -sj index to www and then update or update www and then move the new -sj technology to the index (or something to that effect). It seems like they chose the former. What I'm seeing is the absolute same results as what was/is on -fi for the last two days. New links still aren't in there, but thats the part GG said could happen rather quickly.

[wackmaster]

Does everyone/anyone agree with this:

-sj: Old index, no cached sites to view

-fi: New index, cached sites there

both -sj and -fi: missing serious backlinks, PR calc's

[littlecloud]

This is exactly what GG said would take place. The update would go live on www. first without all the backlinks from April , then the links would be added over a period of time. (hopefully a very short one)

[Need More Hits]

Anyone else getting a 404 on www2
In the last few days I have seen this a couple of times

Don’t know if it is significant or not?

[g1smd]

Hmm, did anyone notice this before?

I did a couple of searches on www that bring up some blogs in the top few positions. Repeating the search on -sj those blogs are nowhere to be seen.

Someone else want to try it?

[merlin30]

I don't see how -fi could be described as a new index. It isn't just that some backlinks are missing; those backlinks that are there represent an old link structure, perhaps 2 months old. I know this is definitely the case as the link references listed in -fi to my site are from the wrong pages - they were the correct pages about 2 months ago but since then some of my link partners have moved my link to other pages. The current index (www) shows the correct backlink structure; -fi (and -sj) have an old structure.

[wackmaster]

Merlin30,

Good, agreed...so:

-sj: Old index, no cached sites to view

-fi: Old index, with some results from Freshie, cached sites there

both -sj and -fi: missing serious backlinks, PR calc's

That being the case, we're looking at - at best - old index results on -sj and -fi, with old backlink structures, and old/screwy PR calc's...

So does anybody still think Google will go live with that? Anything is possible, but sure seems doubtful...

[Birdman]

I just noticed something strange. I see new results when I search from my Opera browser and old results using Toolbar on IE.?

Anyone seen similar?

[Krapulator]

Im seeing -fi results in www. I bloody well hope this is load sharing and not migration!

[Krapulator]

False alarm! www results are back again

[merlin30]

Birdman,

Are your differing results from different browsers a result of DNS cacheing? Your Opera browser makes a DNS query, hits say -sj and caches the result - for the Time To Live on the DNS record you will see -sj results. IE will do a similar thing but has probably cached a different date centre.

[edited by: merlin30 at 11:41 pm (utc) on May 12, 2003]

[BigDave]

merlin30,

That's correct. I've seen the same thing in previous updates and checked the requests with etherial. Different lookups get you different data centers.

[merlin30]

For the duration of this 'update' (nearly 2 weeks now) the traffic to my site has held steady and I know a lot of it comes from Google. If the -fi/-sj index had been propogating through the data centres then I would have expected my traffic to dip as more and more requests resolved to a data centre with the new index. This does not yet seem to be the case.

Here is a point I raised in an earlier thread (but unanswered). I have done lots of pings today on www.google.com and found about 6 different IP addresses; none of these IP addresses relate to the IP addresses of the date centres referred to on sites like google-dance. The question is does anyone know how many data centres actually make up www, and are -sj and -fi included in that set?

[Anon27]

SJ just went to EX, was there for 2-3 minutes then went back to what it was...

[edited by: Anon27 at 12:38 am (utc) on May 13, 2003]

[jojojo]

"SJ and FI just went to EX"

No they didn't.

Everything is the same as it has been.
Update has not started yet.

[coosblues]

i'm not seeing any migration to www(1). The only datacenters which seem the same at the moment are -ex, -sj, and -fi. As of today the only difference I notice is the migration to -ex. I don't see this as an update - just migration changes. Those of you seeing changes on www(1) must be pulling from a different datacenter. Until all datacenters are the same I'd say NO UPdate.