httpd-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From <>
Subject [users@httpd] tracking user clicks thru a session id - G Spillman
Date Tue, 19 Apr 2005 12:23:29 GMT

I'd definately appreciate getting an inside view into what you have

NB Please cc to as this email addy is getting
flooded with rubbish, so i'd probably miss your reply again, like below !

>From : graham.rj.spillman - graham.rj.spillman@<protected>
Date : 2005/04/18 12:00:47


I have written a log analyser in VBA which runs in Excel
Couldn't persuade Analog to do it for me with the results I want.

My need was to focus on response time, so I've added %T to a custom log
formatas below
   LogFormat "%h %t %T \"%r\" %>s %b" performance

My code identifies individual users as it scans the access log
I use IP address to identify users.
You may need to use your SSL string instead - in principle no difference.

It would not be difficult to modify the code in the IPAddress collection to
record what you are looking for and then spit this out on a spreadsheet on
change of day or reaching access log EOF.

I already produce for each page requested the following
(which thereby intrinsically includes the count of D (#Hits) which you need):
   Min response, Median response, Mean response, Max response
   ditto for Bytes returned
   + the page ID requested

There are filtering facilities to get rid of requests for GIFs, Java status
requests etc

... it's all documented too  :o)

I scan up to 10,000 log lines/sec
Perl would do it faster no doubt, but Windoze Excel is good enough for me

Post a reply if you're interested further


-----Original Message-----
From: pete@<protected> [mailto:pete@<protected>]
Sent: 14 April 2005 10:04
To: users@<protected>
Subject: [users@<protected>] tracking user clicks thru a session id


I've got an app where users login, go thru pages A to B, to C, and finish
at D. I'm dumping out in each access log entry the users' SSL session ID
to identify a respective user, e.g. - - [13/Apr/2005:13:20:11 +0000] "POST /PAGEA HTTP/1.1"
70007 544 0BF4C6E30ACRA36FC6016B832697AEB1942B2154068C801ED3F9068450FE9B06 - - [13/Apr/2005:13:21:49 +0000] "POST /PAGEB HTTP/1.1"
70007 544 0BF4C6E30ACRA36FC6016B832697AEB1942B2154068C801ED3F9068450FE9B06

What I want to so is find out is how many users get to each point in the
application, basically so I can report on people getting say to the
checkout (C) but not actually purchasing (D), and the number purchasing

People can cycle thru certain pages,

i.e. login - A - B - A - B - C - D

so I don't really want to just do a "cat access.log | grep 'PAGEA' | wc -l"

Before I get into loads of Perl, does any1 know of a web analysis tool /
script that allows you to say track user clicks thru a series of
predefined pages (A, B etc) and report on where people dropped out ? They
all do lots of reports, but I can't find one that does this....
Any suggestions are appreciated.


This email was sent by
   " - Internet Solutions for the UK"

The official User-To-User support forum of the Apache HTTP Server Project.
See <URL:> for more info.
To unsubscribe, e-mail:
   "   from the digest:
For additional commands, e-mail:

View raw message