hep-puppet / nagios

Puppet module to set up Nagios monitoring for your infrastructure

Puppet Nagios module

This module is based on the internal puppet module of the University of Bristol IT services. Many thanks to @djjudas21 and Paul for providing the source and documentation.

This module to sets up Nagios monitoring for your infrastructure which includes:

• core: domain, CPU, daemons, I/O, networker, memory usage, load, uptime, yum, selinux, certificate, smtp, swap, zombies, users, cron, NTP, TCP traffic, openmanage, xinetd, DNS
• bind: named status
• dhcp: dhcpd status, pools, failover
• mysql: Queries Per Second (QPS), Multi-Master replication Manager (MMM), DB size,
• nagios client: config, stats
• nagios server: config, apache, AAAA record, NRPE
• puppet: last run, environment, master, DB

Usage

Contribution

Please read CONTRIBUTING.md.

nagios's Issues

Change warning threshold for CVMFS

Missing dependencies

A bunch of dependencies are still missing from the private module. The corresponding tests, implementation and documentation need to be added.

Error: /Stage[main]/Nagios::Config::Client/File[/etc/nagios/nrpe.cfg]: Could not evaluate: Could not retrieve information from environment development source(s) puppet:///modules/nagios/nrpe/nrpe.cfg

Error: /Stage[main]/Nagios::Plugins::Core/Nagios::Plugin[check_kernel_passive]/File[/usr/lib64/nagios/plugins/check_kernel_passive]: Could not evaluate: Could not retrieve information from environment development source(s) puppet:///modules/nagios/plugins/check_kernel_passive

Warning: /Stage[main]/Nagios::Cron::Kernel_passive/Cron[kernel-passive-reboot]: Skipping because of failed dependencies

Error: /Stage[main]/Nagios::Plugins::Core/File[/usr/lib64/nagios/plugins/check_hardware_spec]: Could not evaluate: Could not retrieve information from environment development source(s) puppet:///modules/nagios/plugins/check_hardware_spec

Error: /Stage[main]/Nagios::Plugins::Puppet/Nagios::Plugin[check_puppet_checkin_db]/File[/usr/lib64/nagios/plugins/check_puppet_checkin_db]: Could not evaluate: Could not retrieve information from environment development source(s) puppet:///modules/nagios/plugins/check_puppet_checkin_db

and the missing treatment of the global facter variable monitoring_server

Create examples for useful hostgroups

• Computing Element
• worker node
• BDII
• etc

Make 'logged users warning' configurable

vmstat does not return an average for last line

This script purports to check cpu state over a specified interval, but uses vmstat to check state, then assumes the last line gives an average. Suggest changing from vmstat to sar. To overcome the problem of processing float output in bash, can use bc to process.

Linux avg cpu over interval:

  Linux)
    CPU_REPORT=`/usr/bin -u $INTERVAL_SEC $NUM_REPORT | tail -1`
    CPU_USER=`echo $CPU_REPORT | awk '{ print $3 }'`
    CPU_SYSTEM=`echo $CPU_REPORT | awk '{ print $5 }'`
    CPU_IOWAIT=`echo $CPU_REPORT | awk '{ print $6 }'`
    CPU_IDLE=`echo $CPU_REPORT | awk '{ print $8 }'`

Floating point comparison for bash/bc

float_ge() {
  [ $(echo "$1 >= $2" | /usr/bin/bc) -eq 1 ]
}

Floating point comparison in use:

if float_ge "${CPU_IOWAIT}" "${IO_CPU_C}" || float_ge "${CPU_USER}" "${U_CPU_C}" || float_ge "${CPU_SYSTEM}" "${S_CPU_C}";
then
...

Requires packages sysstat and bc to fully implement