__/ [Stacey] on Friday 30 December 2005 17:35 \__
> "John Bokma" <john@xxxxxxxxxxxxxxx> wrote in message
>> Roy Schestowitz <newsgroups@xxxxxxxxxxxxxxx> wrote:
>>> __/ [John Bokma] on Thursday 29 December 2005 22:13 \__
>>>> "kevboy" <themckinneys@xxxxxxxxx> wrote:
>>>>> This seems like a stupid question, but how do I read me referral
>>>>> logs? I have a blog with Blogger, and I'm guessing I should be able
>>>>> to read it, but every time I do a search for how to read them I get
>>>>> a billion hits that just mention the phrase....
>>>>> I'm sure it's very easy, so I apologize if it's a stupid question.
>>>> FTP to your server and download the file access_log
>>> Over 80% of the referrals I see are pure spam,
>> Are you sure about that? I see hardly spam, and if I see it, ah, well, I
>> can do referer spam as well >:->.
>>> so don't get your hopes
>>> high. Inspection of referral logs can be an utter waste of time. This
>>> pattern begins to emerge and gets worse as a site matures, it seems.
>> But 80% sounds crazy, and if that's the case something is very, very
>> I have no idea how much I get, I see now and then spam. Sometimes it
>> sticks out because I get 40 hits in a short time, and hence it ends up in
>> my top 100 (or so) of hits via direct links. Sometimes its quite subtile,
>> 2-3 hits which get swamped in the direct hits/google hits.
> I am not seeing 80% either. I have viewed my raw logs against both the
> webalizer and modlogan. It seems pretty good. I get referrer spam every now
> and then but not bad like 80% maybe around 10 - 15%.
Here are examples from my main domain. I took screenshots from the top and an
arbitrary bit, not adhering to any pattern that supports my argument.
I already have some protection from zombies. AWStats indicates that 1000-2000
continue to attack every day so in absence of protection, the figures would
more like this ( http://www.schestowitz.com/Weblog_Frames/referrer_spam.png
). AWStats manages to set these aside. It's somewhat of a distributed denial
of service attack because each of these zombies requests several megabytes
of data repeatedly. As a consequence, certain bits of the site remain
blocked, even to genuine visitors and search engines. Annoying.