check_uptime / check_snmp_uptime

Documentation is in plugin code as usual for all my plugins.
Here are parts of it:


The plugin can either retrieve information from local system (when you run it through check_nrpe for example) or by SNMP from remote system.

On local system it will execute standard unix 'uptime' and 'uname -a'.

On a remote system it'll retrieve data from sysSystem for system type and use that to decide if further data should be retrieved from
sysUptime (OID 1.3.6.1.2.1.1.3.0) for windows or
hostUptime (OID 1.3.6.1.2.1.25.1.1.0) for unix system or
snmpEngineTime (OID 1.3.6.1.6.3.10.2.1.3) for cisco switches

For information on available options please execute it with --help i.e:
check_uptime.pl --help

As I dont have time for extensive documentation below is all very brief:

1. You can also specify warning and critical thresholds which will give warning or critical alert if system has been up for lees then specified number of minutes. Example:
check_uptime.pl -w 5
Will give warning alert if system has been up for less then 5 minutes

2. For performance data results you can use '-f' option which will give total number of minutes the system has been up.

3. A special case is use of performance to feed data from previous run back into the plugin. This is used to cache results about what type of system it is (you can also directly specify this with -T option) and also means -w and -c threshold values are ignored and instead plugin will issue ONE alert (warning or critical) if system uptime changes from highier value to lower

EXAMPLES =======================================

1. Local server (use with NRPE or on nagios host), warning on < 5 minutes:

define command {
command_name check_uptime
command_line $USER1$/check_uptime.pl -f -w 5
}

2. Local server (use with NRPE or on nagios host),
one critical alert on reboot:

define command {
command_name check_uptime
command_line $USER1$/check_uptime.pl -f -c -P "SERVICEPERFDATA$"
}

3. Remote server SNMP v2, one warning alert on reboot, autodetect and cache type of server:

define command {
command_name check_snmp_uptime_v2
command_line $USER1$/check_uptime.pl -2 -f -w -H $HOSTADDRESS$ -C $_HOSTSNMP_COMMUNITY$ -P "$SERVICEPERFDATA$"
}

4. Remote server SNMP v3, rest as above

define command {
command_name check_snmp_uptime_v3
command_line $USER1$/check_uptime.pl -f -w -H $HOSTADDRESS$ -l $_HOSTSNMP_V3_USER$ -x $_HOSTSNMP_V3_AUTH$ -X $_HOSTSNMP_V3_PRIV$ -L sha,aes -P "$SERVICEPERFDATA$"
}

5. Example of service definition using above

define service{
use std-service
hostgroup_name all_snmp_hosts
service_description SNMP Uptime
max_check_attempts 1
check_command check_snmp_uptime
}

6. And this is optional dependency definition for above which makes every SNMP service (service beloning to SNMP servicegroup) on same host dependent on this SNMP Uptime check. Then if SNMP daemon goes down you only receive one alert

define servicedependency{
service_description SNMP Uptime
dependent_servicegroup_name snmp
}