Configuring Alert Rules
Use the following configuration to configure the alert rules.
configure
alerts rules group alert_group_name
interval-seconds seconds
rule rule_name
expression promql_expression
duration duration
severity severity_level
type alert_type
annotation annotation_name
value annotation_value
exit
exit
NOTES:
-
alerts rules – Specifies the Prometheus alerting rules.
-
interval-seconds seconds – Specifies the evaluation interval of the rule group in seconds.
-
group alert_group_name – Specifies the Prometheus alerting rule group. One alert group can have multiple list of rules. alert_group_name is the name of the alert group. The alert-group-name must be a string in the range of 0 through 64 characters.
-
rule rule_name – Specifies the alerting rule definition. rule_name is the name of the rule.
-
expression promql_expression – Specifies the PromQL alerting rule expression. promql_expression is the alert rule query expressed in PromQL syntax. The promql_expression must be a string.
-
duration duration – Specifies the duration of a true condition before it is considered true. duration is the time interval before the alert is fired.
-
severity severity_level – Specifies the relative level of urgency for the operator's attention. severity_level is the severity level of the alert. The severity levels are: critical, major, minor and warning.
-
type alert_type – Specifies the type of the alert. alert_type is the user-defined alert types. For example, Communications Alarm, Environmental Alarm, Equipment Alarm, Indeterminate Integrity Violation, Operational Violation, Physical Violation, Processing Error Alarm, Quality of Service Alarm, Security Service, Mechanism Violation, or Time Domain Violation.
-
annotation annotation_name – Specifies the annotation to attach to the alerts. annotation_name is the name of the annotation.
-
value annotation_value – Specifies the annotation value. annotation_value is the value of the annotation.
The following example monitors the success rate of SMF session creation by configuring Prometheus alert rule to report if session creation is less than threshold.
cee# configure terminal
alerts rules group SMFProcStatus
interval-seconds 300
rule PDNSessCreate
expression "sum(increase(smf_service_stats{app_name=\"SMF\",procedure_type=\"pdn_sess_create\",status=\"success\"}[5m])) / sum(increase(smf_service_stats{app_name=\"SMF\",procedure_type=\"pdn_sess_create\",status=\"attempted\"}[5m])) < 0.95"
severity major
type "Communications Alarm"
annotation summary
value "This alert is fired when the success percentage of pdn_sess_create procedure is lesser threshold”
exit
In the following example, a alert is sent as SNMP Trap to receiving agent when a snmp-trapper is configured.
cee# configure terminal
snmp-trapper enable true v2c-target 172.16.181.41 community public port 161
exit
The following example configures an alert, which is fired when the percentage of UDM responses is less than the specified threshold limit.
Example:
cee# configure terminal
alerts rules group SMFUDMchk_incr
interval-seconds 300
rule SMFUDMchk_incr
expression "sum(increase(smf_restep_http_msg_total{nf_type=\"udm\", message_direction=\"outbound\", response_status=~\"2..\"}[3m])) / sum(increase(smf_restep_http_msg_total{nf_type=\"udm\", message_direction=\"outbound\"}[3m])) < 0.95"
severity major
type "Communications Alarm"
annotation summary
value "This alert is fired when the percentage of UDM responses is less than threshold“
exit
exit
exit
You can view the configured alert using the show running-config alerts command.
Example:
The following example displays the alerts configured in the running configuration:
cee# show running-config alerts
interval-seconds 300
rule SMFUDMchk_incr
expression "sum(increase(smf_restep_http_msg_total{nf_type=\"udm\", message_direction=\"outbound\", response_status=~\"2..\"}[3m])) / sum(increase(smf_restep_http_msg_total{nf_type=\"udm\", message_direction=\"outbound\"}[3m])) < 0.95"
severity major
type "Communications Alarm"
annotation summary
value "This alert is fired when the percentage of UDM responses is less than threshold“
exit
exit
exit