Nagios/icinga plugins to monitor an openstack installation.
Find also some information about these plugins in the talk Monitoring an Openstack Cluster with icinga/nagios held at the 11th Swiss OpenStack User Group Meetup by Cirrax.
For all checks there are some common arguments:
-h, --help show help message and all arguments
--filename FILENAME File to read openstack credentials from. If not set it
takes the standard environment variables from openstack.
-v, --verbose increase output verbosity (use up to 3 times)(not
everywhere implemented)
--cacert CACERT Specify a CA bundle file to use in verifying a
TLS(https) server certificate.
--insecure Explicitly allow to perform "insecure" SSL
(https) requests. The server's certificate will not be
verified against any certificate authorities. This
option should be used with caution.
--timeout TIMEOUT amount of seconds until execution stops with unknow
state (default 10 seconds)
Currently the following checks are implemented:
Nagios/Icinga plugin to check running cinder agents/services.
This corresponds to the output of 'cinder service-list'.
optional arguments:
-w RANGE, --warn RANGE
return warning if number of up agents is outside RANGE
(default: 0:, never warn)
-c RANGE, --critical RANGE
return critical if number of up agents is outside
RANGE (default 1:, never critical)
--warn_disabled RANGE
return warning if number of disabled agents is outside
RANGE (default: @1:, warn if any disabled agents
--critical_disabled RANGE
return critical if number of disabled agents is
outside RANGE (default: 0:, never critical
--warn_down RANGE return warning if number of down agents is outside
RANGE (default: 0:, never warn)
--critical_down RANGE
return critical if number of down agents is outside
RANGE (default: 0, always critical if any
--binary BINARY filter agent binary
--host HOST filter hostname
Admin rights are necessary to run this check.
Nagios/Icinga plugin to check running neutron agents. This corresponds to the output of 'neutron agent-list'.
optional arguments:
-w RANGE, --warn RANGE
return warning if number of up agents is outside RANGE
(default: 0:, never warn)
-c RANGE, --critical RANGE
return critical if number of up agents is outside
RANGE (default 1:, never critical)
--warn_disabled RANGE
return warning if number of disabled agents is outside
RANGE (default: @1:, warn if any disabled agents
--critical_disabled RANGE
return critical if number of disabled agents is
outside RANGE (default: 0:, never critical
--warn_down RANGE return warning if number of down agents is outside
RANGE (default: 0:, never warn)
--critical_down RANGE
return critical if number of down agents is outside
RANGE (default: 0, always critical if any
--binary BINARY filter agent binary
--host HOST filter hostname
Admin rights are necessary to run this check.
Nagios/Icinga plugin to check floating ip's. Counts the assigned ip's (= used + unused).
This corresponds to the output of 'neutron floatingip-list'.
optional arguments:
-w RANGE, --warn RANGE
return warning if number of assigned floating ip's is
outside range (default: 0:200, warn if more than 200
are used)
-c RANGE, --critical RANGE
return critical if number of assigned floating ip's is
outside RANGE (default 0:230, critical if more than
230 are used)
Admin rights are necessary to run this check.
Nagios/Icinga plugin to check running nova services.
This corresponds to the output of 'nova service-list'.
optional arguments:
-w RANGE, --warn RANGE
return warning if number of up agents is outside RANGE
(default: 0:, never warn)
-c RANGE, --critical RANGE
return critical if number of up agents is outside
RANGE (default 1:, never critical)
--warn_disabled RANGE
return warning if number of disabled agents is outside
RANGE (default: @1:, warn if any disabled agents
--critical_disabled RANGE
return critical if number of disabled agents is
outside RANGE (default: 0:, never critical
--warn_down RANGE return warning if number of down agents is outside
RANGE (default: 0:, never warn)
--critical_down RANGE
return critical if number of down agents is outside
RANGE (default: 0, always critical if any
--binary BINARY filter agent binary
--host HOST filter hostname
Admin rights are necessary to run this check.
Nagios/Icinga plugin to check nova hypervisors.
This corresponds to the output of 'nova hypervisor-stats'
optional arguments:
-H HOST, --host HOST hostname where the hypervisor is running if not
defined (default), summary of all hosts is used
-w RANGE, --warn RANGE
return warning if number of running vms is outside
RANGE (default: 0:, never warn)
-c RANGE, --critical RANGE
return critical if number of running vms is outside
RANGE (default 0:, never critical)
--warn_memory RANGE return warning if number of disabled agents is outside
RANGE (default: 0:, never warn
--critical_memory RANGE
return critical if number of disabled agents is
outside RANGE (default: 0:, never critical
--warn_memory_percent RANGE
return warning if number of disabled agents is outside
percent RANGE (default: 0:90, warn if 90% of memory is
used
--critical_memory_percent RANGE
return critical if number of disabled agents is
outside percent RANGE (default: 0:90, critical if 95%
of memory is used
--warn_vcpus RANGE return warning if number of down agents is outside
RANGE (default: 0:, never warn)
--critical_vcpus RANGE
return critical if number of down agents is outside
RANGE (default: 0, always critical if any
--warn_vcpus_percent RANGE
return warning if number of down agents is outside
RANGE (default: 0:90, warn if 90% of vcpus are used)
--critical_vcpus_percent RANGE
return critical if number of down agents is outside
RANGE (default: 0:95, critical if 95% of vcpus are
used
Admin rights are necessary to run this check.
Nagios/Icinga plugin to check ceilometer statistics. Returns the statistic of the chosen meter. This also returns the age of the last sample used to aggregate. So this check can also be used to verify freshness of samples in the ceilometer DB. (or of course to check the value).
-m METER_NAME, --meter METER_NAME
meter name (required)
-t VALUE, --tframe VALUE
Time frame to look back in minutes
--tzone TZONE Timezone to use. Ceilometer does not store any
timezone information with the samples.
-w RANGE, --warn RANGE
return warning if value is outside RANGE (default: 0:,
never warn)
-c RANGE, --critical RANGE
return critical if value is outside RANGE (default 0:,
never critical)
--warn_count RANGE return warning if the number of samples is outside
RANGE (default: 0:, never warn
--critical_count RANGE
return critical if the number of samples is outside
RANGE (default: 0:, never critical
--warn_age RANGE return warning if the age in minutes of the last value
is outside RANGE (default: 0:30, warn if older than 30
minutes
--critical_age RANGE return critical if the age in minutes of the last
value is outside RANGE (default: 0:60, critical if
older than 1 hour
--aggregate AGGREGATE
Aggregate function to use. Can be one of avg or sum
(avg is the default)
Nagios/Icinga plugin to check keystone. The check will get a token and mesure the time used.
-w RANGE, --warn RANGE
return warning if number of up agents is outside RANGE
(default: 0:, never warn)
-c RANGE, --critical RANGE
return critical if number of up agents is outside
RANGE (default 1:, never critical)
Nagios/Icinga plugin to check rally results. Takes the outpup of 'rally task results' as input on stdin. and calculates the sum of load- and full- duration and the number of failed scenarios.