On Wednesday, July 12, 2006 12:53 PM [EDT], Gal Ori Steinitz <[EMAIL PROTECTED]> wrote:
Is it possible to get a "unique downloads" number that would be calculated by taking the unique combinations of IP + files ? I am seeing multiple requests for the same file by the same host, because of new podcast clients that download the files in small chunks. (Sometimes as many as 10,000 separate records for the same file). Needless to say this is skewing my reporting and now I can't get a real number anymore representing the number of files people have downloaded. By importing the raw log into a database and doing a select distinct on those two columns I was able to get a real number again. Can Analog do something similar ?
No. This is a version of FAQ 128: http://analog.cx/docs/faq.html#faq128
The podcast clients that cause this register a status "206" instead of "200" in the log. I saw on the Analog mailing list archives a reference to a similar situation with a PDF file - however in that case each download was preceded by a single status "200", therefore it was possible to simply exclude the "206" and get a real number. In my case, these podcast clients only register status "206" for all requests, and never "200".
Are you logging referrers? Is the referrer for the first part different from the other parts? If your web server can create a "sessionID" that can be logged, that might be a solution too. If you are using Apache, you might be able to create a custom logformat that includes an entry that concatenates the IP address and filename together.
Aengus
+------------------------------------------------------------------------ | TO UNSUBSCRIBE from this list: | http://lists.meer.net/mailman/listinfo/analog-help | | Analog Documentation: http://analog.cx/docs/Readme.html | List archives: http://www.analog.cx/docs/mailing.html#listarchives | Usenet version: news://news.gmane.org/gmane.comp.web.analog.general +------------------------------------------------------------------------
