Inventor wrote:
While enthusiastically writing my new program, I was having so much
fun coding and reading www.perl.com articles that I didn't notice a
slow, creeping horror that was gradually building up on my machine.
Luckily I was reading about Apache and the log files when I remembered
that someone here advised me to check the error logs as standard
practice. As I paged through the file using the more command, I
noticed the little 0% indicator never got bigger than 0% for some
reason. Suddenly realizing my peril i held the spacebar down and
reams of text scrolled by but that 0% never went up! A quick ls -l
revealed that the error file had been slowly growing to as size of 2.2
gigabytes during Apache's first week of running! In my newbiness I
did not realize that all those undeclared variables in a number
crunching program were creating error messages from Hades that were
eating up my little hard drive fast! Fortunately the problem was
caught in time and easily fixed. It only took a few hours to delete
the error file and clean up the code, and now the program runs much
faster as well. Whew! I hope you got a chuckle out of that little
adventure. Now to my questions.
I don't know much about Apache, just what I read in the documentation.
Apache, by default, rolls over its error.log and access.log once a
month. That means you can go back 11 months. January's log replaces
last year's January log, etc. You can set it to roll over weekly, or
even daily. I can't tell you the exact command to do this; you'll have
to look it up yourself.
My program accesses small text configuration files, allowing the user
to change them with a CGI interface before writing them out and doing
its compute thing. The problem is that each user needs to have their
own set of configuration files so that the users don't all step on
each other's configuration. I would like for each user to be able to
return again and again and always find their configuration files still
on the server, eagerly awaiting their chance to be accessed yet
again.
Cookies are the answer.
I have worked on web sites, that for political reasons, can not use
cookies. They're a pain to maintain since you have to give every
"session" a unique ID. The problem is that old IDs get cached (and
sometimes re-used) and then you get complaints that you site is not
working properly. Or worst, it sends info to the wrong place.
I imagine I need some sort of authentication system, but the prospect
of writing one seems daunting. Are there any modules for doing that?
I looked on CPAN and didn't find much of anything obvious.
I haven't done this but the best method is to use https rather than the
old method of htaccess.
Should I make copies of the files with username.config.txt or
something like that?
What is the best way to do the copying in Perl?
See `perldoc File::Copy`.
Do you have any thoughts or suggestions about how to set this up
properly? I'm kind of uncertain about how to proceed.
Set up a local Apache for: development and test. Set up Apache for beta
and production. Set up automatic bug reporting for test and beta.
Thanks in advance.
--
Just my 0.00000002 million dollars worth,
Shawn
"For the things we have to learn before we can do them, we learn by
doing them."
Aristotle
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
http://learn.perl.org/
