#######################################################################
# Sample configuration similar to http://www.linux.cz/stats/mrtg-rrd/ #
#######################################################################

All files here are written by and (c) Jan "Yenya" Kasprzak
<kas@fi.muni.cz>. The files are distributable under the terms of GNU
General Public License, version 2 (not any later version).

This is a set of files used for generating the server statistics
for the ftp/www.linux.cz server. Statistics are available at
http://www.linux.cz/stats/mrtg-rrd/ and are generated by mrtg-rrd.cgi
script (http://www.fi.muni.cz/~kas/mrtg-rrd/).

These files are provided as is - do not bother to send any bugreports.

A brief description:

mrtg-fast.cfg
mrtg.cfg
	these two files are included in mrtg-rrd.cgi and describe all the
	statistics that are measured on the server. The -fast variant contains
	a definition ofn "eth0" interface, which is gigabit interface and
	the counter can overflow in a half minute on 32-bit system. So
	I run "mrtg ./mrtg-fast.cfg" every minute, and "mrtg ./mrtg.cfg"
	every 5 minutes from cron (now I have a real 64-bit system, so
	the overflow problem is only historical)
explore_* scripts
	They are run by mrtg and they generate various statistics. The output
	is conformant to the "Target[]: `...`" syntax of mrtg.
explore_cpu
	Displays non-idle and system time of all CPUs (from /proc/stat).
	For non-idle it sums all the times except iowait, and for system
	it sums system + irq + softirq time ticks.
explore_fan
	Reads the fan RPM value from sysfs.
explore_ftpd
	Parses the output of "ftpcount" command.
explore_load
	Returns the number of processes running (ps ax|wc -l) and 5-minute
	load average (/proc/loadavg).
explore_mailq
	Warning: misleading name. It counts the number of active outgoing
	SMTP connections of Qmail (it has limit of 120 connections, so from
	this graph I can see whether delayed deliveries can be expected).
	Qmail-specific parsing of last few lines in /var/log/maillog.
explore_smart
	The crappy parsing of smartctl output and displaying the temperature
	value. In 2.6 it is unusable because it needs root access.
explore_temp
	This reads the temperature status from system sensors.
explore_vm
	This prints the size of Active and Inactive VM lists, so that you
	can detect kernel memory leak or an overall VM pressure.
hdd.snmp
	Parses /proc/diskstats or /proc/partitions and prints the sectors
	read/written and the drive busy time (i.e. time for which the drive
	had at least one command in flight).
snmpd.conf.add
	Add this to your snmpd.conf to allow hdd.snmp be called from snmpd.


