> On Thu, 26 May 2005 [EMAIL PROTECTED] wrote: > > > Hello, > > > > I've been running Analog for a few months now and I came across a funny > > thing showing up in the report. Analog analyzes all data in the Apache > > log, including file size. The last analasys is from 17-03-2005 to > > 26-05-2005 (70 days). And that makes it even more obvious that something > > is not right. > > > > The statistics of this morning said the total data transferred was 4.01 > > Gigabytes. However when I checked the page again in the evening, it > > suddenly said 1.18 Gigabytes. Then I started figuring and calculating, and > > I figured that it crosses the 4GB data transferred 2 times per month. Why > > does analog "reset" the counter after 4GB of data? is it unable to > > calculate any further? > > > > How can I fix this problem? or is it a bug? > > I run Analog 6.0 on Windows 2000AS on the Apache logs. > > > > Plausible guess, but in fact wrong. There is no such limit. (The number of > bytes is stored in a double, not an unsigned long). You can try it out with > this logfile: > > statslab.cam.ac.uk - - [31/Dec/1999:10:11:12 +0000] "GET /sample.html > HTTP/1.0" 200 3234000000 "http://referrer.com/"; "Mozilla/4.0 (compatible; > MSIE 4.01; Windows 98)" > statslab.cam.ac.uk - - [31/Dec/1999:11:11:12 +0000] "GET /sample.html > HTTP/1.0" 200 3234000000 "http://referrer.com/"; "Mozilla/4.0 (compatible; > MSIE 4.01; Windows 98)" > > You should see 6.02 GB transferred. > > So what is going wrong? Maybe your logs are being rotated, so you're losing > some old logs from the analysis periodically? > > -- > Stephen Turner, Cambridge, UK http://homepage.ntlworld.com/adelie/stephen/ > "Low Priced Cambridge Clare College. Big selection at eBay UK!" > (Ad after Google search for Clare College Cambridge)
I don't have a clue where it goes wrong then. I don't rotate log-files, I do have multiple log-files that are read in the same report, but that's just to make the virtual hosts (vhost.domain.nl) to be logged as domain.nl/vhost However, since all the logs are in an identical format, I don't suspect this is the case. And if I'd be loosing logs periodically, I suppose the analasys would not be the 70 days that it stands for. I am willing to provide you with a link to my Analog's output if that is usefull? I haven't manually calculated what it should be in total, since the logfile's are around 20 Megabytes in size. Could it be that analog 'randomly' picks a log-file for which it's total size is calculated? there is 1 log that is quite small because that site is fairly quiet. But that site never even gets close to 1GB of data. the rest of the sites are all multiple GB's of traffic by now. Any clue's? (Thanks for the quick reply btw!) +------------------------------------------------------------------------ | TO UNSUBSCRIBE from this list: | http://lists.meer.net/mailman/listinfo/analog-help | | Usenet version: news://news.gmane.org/gmane.comp.web.analog.general | List archives: http://www.analog.cx/docs/mailing.html#listarchives +------------------------------------------------------------------------
