You probably see stats for your sites which have the following:
- Total number of visitors
- Total number of unique visitors
- # of new visitors
- # of return visitors
My questions is how they know/store this information? Consider storing all IPs is very inefficient. Do you agree?
Any suggestion/reply is great! Thanks
It depends on the stats program they use, but what it comes down to is something processes your Web server's raw log files to produce those results. Whether they preprocess them and store the results in a database (saving disk space), or keep copies of the Web server logs (with every single request) and process them whenever you view the stats depends on the program being used.
Thanks for lightning responses of Freddy and Kevin.
I read the article Freddy pointed and did learn some interesting stuff.
Process the data when being requested in this situation won't be possible for huge amount of data. So the solution will be: preprocessing them at some certain time during the day.
However, basically there is no way we could process these information without storing IPs from the past?
But what if the period is a week or even a month? In that case, I believe that adding up one day's data is not correct (correct if I'm wrong). And storing data for such a period takes tons of space!
Well I stepped into this one quite late gentlemen, not to mention after several large asahi beers, so I will bow out now and return to see what has transpired in the morning, you never know though my cat Freddy might be on later cheers!