|Google IP with invalid User-agent|
As I am always analyzing my raw logs, I noticed the following entry:
184.108.40.206 - - [14/Nov/2005:14:13:03 +0100] "GET /foo/bar/ HTTP/1.0" 200 4362 "-" "Mozilla/4.0 (compatible; MSIE 6.0;)" "212.141.2.*, unknown"
A member of the german ABAKUS SEO Forum reported similar sightings of this invalid User-agent from a Google IP, three days after mine. This is why I am posting it here.
What is it? A human Google quality rater? Or only someone using Google Accelerator?
sidenote: After some days, traffic on the site (existing since 2001, >1000 pages unique content, more than 6 ODP listings, 2 non-paid Yahoo Directory listings, 2 VLIB listings, misc. college links) visited by the Google IP started to drop to about 50%. Until today.
I do not believe this is a coincidence.
The IP address you mention might be used by Google WAP or CHTML proxy, although most accesses from these proxies on my site are comming from the Google IP 220.127.116.11.
maybe it's testing if you're clocking? i'm nore sure if google use ip in the same netmask as crawler
Could it be related to the Google translation? Do a Google search and you will see traffic logs with this line:
18.104.22.168Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1),gzip(gfe) (via translate.google.com)
Google human raters don't usually fake their UA. They use Firefox ; ) I can give you the ip addresses for both USA and UK if you want.
And while we're talking about IP addresses I'd like to know what's going on with 66-194-6-xx.gen.twtelecom.net - a lot of visits to penalized sites with no referrer.
I am a little indignant about the resposes. Having enough experience in log analysis to distinguish a cell phone proxy, translate.google.com and Google service related stuff in general from other log entries, this one was posted here, because I was not the only one who had that visit. There is a minor chance that it was only a Google accelerator user. I initially hoped to find others with similar experiences and not half-baked guesses. Sorry for the rant.
|Having enough experience in log analysis to distinguish a cell phone proxy, translate.google.com and Google service related stuff in general from other log entries, this one was posted here, because I was not the only one who had that visit. |
First of all, I found the IP address you mentioned also in my log files and my answer in msg #2 was based on my analysis of that visit. The visitor surfed my website like a normal user: the time it spent on each page was comparable with the amount of content per page, i.e. larger pages took more time before a jump to the next page. There was also a logical sequence in the fetched pages from general information to more specific. It is highly unlikely that a bot would mimic a human in such a way, so I strongly believe that it was a human.
Secondly the visitor read my Dutch language pages, although English language pages about the same subject were also available. I remember from a thread some time ago that Google only recruted quality raters from English speaking countries so IMHO it is unlikely that it was a Google quality rater.
Third, I have several entries in my logfile from the Google WAP/CHTML proxy at IP 22.214.171.124 and 126.96.36.199 with exactly the UA string you mention: "Mozilla/4.0 (compatible; MSIE 6.0; )". Therefore it was not a guess, that this IP might be used as a secondary IP for the WAP/CHTML proxy, but based on the content of my logfiles.
Fourth, Google has a very strict way of assigning IP addresses. Look at the numbers for the DC's for example. They all end at the same numbers: 104, 99, etc. Therefore--although without scientific proof--the fact that the two IP addresses I know of the WAP proxy end in 136 is an indication that other addresses ending in 136 are also proxies.