In this chapter we will cover some very basic ways to determine the status of a system by directling you to the log files, cron files, and emergency rescue operation, or "Call out the sheriff, little Jimmy fell in the well again."
I. Syslogd and klogd
B. Klogd intercepts kernel messages and logs them as a part of syslogd in /var/log/messages. The klog entries are relatively easy to see as they contain the word 'kernel' in the prefix of the log line.
II. Crontab and logrotate
B. Cron also scans the /etc/crontab as well as the /var/spool/cron/ directory for user crontab entries. Creating of the /var/spool/cron/<user> crontab is via the 'crontab' utility. Read the man page for further info on creating user crontabs.
C. Logrotate is run as a cron job from the /etc/cron.daily directory. This program will rotate logs on a daily, weekly, monthly or when a file becomes too large. It is configured by the /etc/logrotate.conf file and several specic files in the /etc/logrotate.d/ directory.
III. Emergency boot disk creation
B. Obtain a floppy disk, any 1.44M floppy disk, and place it in the drive. WARNING: This will erase all data!
C. As root, edit the /sbin/mkbootdisk shell script and make the following
changes at about line 117. It should look like this:
D. Type /sbin/mkbootdisk.
E. Test the disk by leaving it in the drive and rebooting the machine.