Re: how to read log report

Simran Singh (Simran.Khalsa@Eng.Sun.COM)
Tue, 27 Jun 1995 10:49:42 +0800

> Reply to: RE>>how to read log report
> jennifer, thanks for all your notes...
> here's my question... if someone opens an html page that has 5 in-line gifs
> that get called up (say, thumbnails), but they've only clicked in one place
> somewhere else to open the page, does that end up reading as 1 file or 6
> files....
> thanks everybody!
> julia

Jennifer basically gave us an accurate reading on the log files (thanks!).
The answer to Julia's next question is 6. The http protocol specifies
one access for each file, which boils down to one for the document
being accessed, and one for each inline image file (or inline sound.
video, etc file when the browsers get around to supporting them :-).

However, some files get cached in the browser. For example, if you
have three pages, each with the same 3 in-line gifs at the bottom,
(used for navigation), then for the first page, the browser will access
the page, then the inline gifs. For the second page, the browser will
access the page, then realize that it already has the gifs, and not
access them again.

If a site uses a caching proxy server (they're getting more and more common,
especially at large sites), then people browsing from that site will
get their pages from the cache on the server, rather than getting the
page from us. (That is, if someone on that site has already accessed
that particular page (or inline gif or etc)). Jennifer mentioned this

Also, there's one teeny correction I'll make to Jennifer's reading:

| >Sites
| >-----
| > 17
| > 8
| > 8
| [etc.]
| This is basically how many files each site accessed.
| That is, viewed 17 files.

Actually it indicates how many accesses each site had. usually, because of
the caching mentioned above, that works out to a one-to-one correspondence
between accesses and files, but in some circumstances (cache refreshes,
multiple people browsing from one site, etc), a site can access the same
file more than once.

And finally, a few more details:

>The Access Log
> - - [26/Jun/1995:04:21:41 -0700] "GET /Poets/Jennifer/mosaic/c9.gif HTTP/1.0" 200 8766
> - - [26/Jun/1995:04:24:25 -0700] "GET /Poets/Jennifer/index.html HTTP/1.0" 200 2270

There's a bit more information hidden in the full line from the access log:
"GET" means the "get" method, i.e. the browser requested the file. This
field might be "POST", which usually corresponds to hitting the
"DO IT" button on a form.
"HTTP/1.0" means the protocol used for the request (or something, I'm not
really sure about this field,
The last number, I believe, is the file size, and the
Second-to-last number escapes me, perhaps it's the browser code, so you can
tell what kind of browser accessed the page? I know that information is
in the protocol, so if you want to get fancy, you can write scripts that
give the viewer a different page based on what kind of browser they use.

Have fun,