Re: [analog-help] Re: Why count status 206?

[William Torrey]

----- Original Message ----- 
From: "Rick Hardison" <[EMAIL PROTECTED]>
To: <analog-help@lists.meer.net>
Sent: Friday, January 14, 2005 5:57 PM
Subject: [analog-help] Re: Why count status 206?


Rick Hardison <[EMAIL PROTECTED]> writes:
By default analog counts status 206 as a distinct successful reference.
However, I have a large file that gets downloaded from my site.
In looking at my access log, I notice that sometimes a download of my 
large
file is complete with one 200 entry.

But other times, the file has multiple entries, 200 being the first entry
followed by multiple 206 entries.
I think this is caused by the browser requesting smaller portions at a 
time
in
order to download the whole file in smaller pieces.
If this is correct, it would seem that I should only count the 200s in 
order
to
get the actual number of distinct download requests for the individual 
file.

But since analog counts 206s by default, I wonder if I am understaind 
this
correctly.

Any insight appreciated.
Thank,
Rick

Thanks to all for the info. I have looked again at my access log and 
noticed
that most of my 206s come immedidately after a 200 from the same ip 
address. I
am not sure what is going on except that the client browser must be asking 
for
smaller (partial) pieces. In any case, if I count 206s I will way 
overcount how
many logical requests for my file were made. I want to count how many 
times
someone requested my file, not how many transfers it took to actually 
download
the file. So I guess I will exclude 206s and then run the risk of missing 
a few
unique 206s. If anyone knows how to have it both ways I would appreciate 
some
info. Thanks again. Rick
Try looking at it from a raw bytes transfered POV for the file you are 
interested in. Then all you need to do is divide the bytes transferred by 
the file size and you have approx # of downloads. Rounding down shoul help 
reduce errors.


+------------------------------------------------------------------------
|  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
+------------------------------------------------------------------------