I looked at several trouble-ticket and user-request systems, and they're dandy for collaboration between help desk people. Unfortunately, they don't really do the trick for organizing my day-to-day information. Here are some of the more interesting packages and papers I saw:
- Bugzilla--server software designed to help manage software development.
- BATTS (Barnhard Associates Trouble Ticketing System)--a Perl system with an excellent breakdown of steps to assign tickets, resolve them, close them, and more; it requires a database
- OTRS (Open Source Ticket Request System)--a nice setup, but it requires me to log in more than once
- Call Center, Bug Tracking and Project Management Tools for Linux
- Bugaboo--a web-based bug reporting system for Linux, written in Pascal with its own database included
- Frontdesk--an older package, but with a basic approach that might work
- Painless Bug Tracking--an excellent paper by Joel Spolsky
- Request Tracker
- Request Tracker source
The one that seems to have improved the most is Roundup. Its features include:
- Simple installation: it takes about 30 minutes and requires no additional software other than the Python interpreter.
- Simple use: accessible through the Web, email, the command line, or Python.
- High configurability: the web interface is fully editable.
- Documentation: for installation, upgrading, maintenance, users, and customization.
- Command-line access: you can script it using standard shell scripting and can use the API from other Python programs.
What Seems to Work So Far
Nothing was quite right, so I invented my own system.
The idea of doing a complete brain dump and storing everything in a trusted location was probably the most useful thing I learned from the Getting Things Done book.
My job as a system administrator doesn't change every day, but it's much easier to keep track of things via date rather than via subject. I tend to remember things in time order, so I finally stopped trying to change the way I work to fit some hierarchy. Instead, I made a directory structure on the machine to match my work habits.
I have a top directory cleverly named notebook with subdirectories in the form yyyy/mmdd, so every day has its own folder (Figure 2).
Figure 2. Every day has its own folder
- On April 11, 2000, I grabbed some informational documents from the Sun web site.
- On December 30, 2000, I found something about setting up /dev/random under Solaris.
- On November 13, 2005, I'll have something that needs attention in my agenda.
Someone might ask me, "Remember that thing we broke last Wednesday, and then we fixed it on Thursday before anyone noticed?" (See Figure 3.)
Figure 3. Finding a file by date
Make sure to index the files you use and modify most often for rapid retrieval. grep is fine for doing a quick-and-dirty search through a few files, but it doesn't work nearly as well for anything larger.