Try process monitoring. Write a script (you choose language) that monitors your process space (ps -ef etc) and reports to you when "unusual events" occur... You define "unusual" for yourself for learning purposes... For example...
1. A process that appears in less than 1 of 4 sweeps
2. Processes that run more than a few minutes
3. Process whose actual time exceeds its cpu time by more than 100 to 1
things like that...
Become a sed expert... Set up a crontab (again an example) and write a sed script that advances the time of every job in the list by one hour...
Become an awk expert -- even better for a sysadmin
There are lots of basic tools you could teach yourself to use proficiently for admin purposes, and you could do useful things with them like learn to automate system backups, etc.
When you feel really confident, have someone come along who knows what he/she is doing (so there's no physical damage) and DESTROY your system -- screw up NFS, over-flow the process space, what ever, and then you have to (a) figure out what is wrong and (b) fix it...
quine@sonic.net