WikiSysop at 20:15, April 7, 2022

2022-04-07T20:15:41Z

Tgilchri: Published

2021-12-22T13:54:03Z

Published

New page

{{ArticlePEServiceObservability
|Monitoring=The metrics described in this document pertain to the internal monitoring of Genesys CX Insights (GCXI) and Reporting and Analytics Aggregates (RAA) performance and containers, and are intended for administrator use only. These metrics are distinct from the metrics which GCXI provides for monitoring contact center activity, which are described in {{Link-SomewhereInThisVersion|manual=RPRT}}

===About monitoring in GCXI===
GCXI container provides TCP metrics endpoint (Prometheus format) on port 8180, and HTTP endpoint at: '''<gcxi_url>/gcxi/monitor/metrics''', for example <nowiki>https://gcxi-22-30.genhtcc.com/gcxi/monitor/metrics</nowiki>.

===About monitoring in RAA===
In RAA, the '''Health port''' retrieves the health status. It can take as much as three minutes for RAA to analyze Health status.

In addition to health metrics, RAA provides other metrics that capture statistical information and other data extracted using the '''Metrics port'''. The first time RAA scrapes this metric data, it aggregates data beginning from the the earliest data available. On subsequent occasions, RAA aggregates the data over the time period that has elapsed since the previous scrape. RAA reads this data from a local file that contains only two or three scrape intervals, which speeds the process; the aggregation of this statistical information typically takes only a few seconds.
|EnableMonitoring=Monitoring is not enabled by default for GCXI or RAA.
===Enable GCXI monitoring===
GCXI Helm Chart ships standard ServiceMonitor object for integration with Prometheus Operator, in [https://git.scm.genesys.com/Genesys/gcxi-docker/blob/dev/k8s/helm/gcxi/templates/gcxi-service-monitor.yaml gcxi-service-monitor.yaml].

Related Helm Chart values:
<source lang="bash">gcxi.deployment.deployServiceMonitor = false|true
</source>whether to deploy ServiceMonitor object for integration with Prometheus Operator
<source lang="bash">gcxi.ports.worker.metrics = 8180 (default)
container port for metrics endpoint</source>
===Enable RAA monitoring===
RAA exposes metrics with the help of a sidecar container.

#To launch the sidecar container, you specify the '''raa.statefulset.containers.monitor''' element (including ports for metrics and/or health endpoints) in '''values.yaml'''. The value of '''raa.env.STAT_SCRAPE_INTERVAL''' must exceed the scrape interval, in seconds, configured in Prometheus for the metrics port.
#:For example:
#:<source lang="bash">
raa: ...
env:
STAT_SCRAPE_INTERVAL: 15
...
statefulset:
...
containers:
...
monitor:
name: "{{$.Chart.Name }}-monitor"
...
metrics:
portName: "metrics"
containerPort: "9100"

health:
portName: "health"
containerPort: "9101"
...
</source>
#If Prometheus Operator is configured in your Kubernetes environment, then you can configure Prometheus using the parameters '''PodMonitor''' and '''PrometheusRule''' in your '''values.yaml''' file. When configuring '''interval''' and '''scrapeTimeout''' for the health port, take into account that health check usually takes around a minute.
#:For example, set the following values:
#:<source lang="bash">
raa:
...
podMonitor:
name : "{{ tpl $.Values.raa.serviceName $ }}-monitor"
podMetricsEndpoints:
- port: "{{ $.Values.raa.statefulset.containers.monitor.metrics.portName }}"
path: "/{{ tpl $.Values.raa.statefulset.containers.monitor.metrics.portName $ }}"
interval: "{{ $.Values.raa.env.STAT_SCRAPE_INTERVAL }}s"
- port: "{{ tpl $.Values.raa.statefulset.containers.monitor.health.portName $ }}"
path: "/{{ tpl $.Values.raa.statefulset.containers.monitor.health.portName $ }}"
interval: 4m
scrapeTimeout: 3m
podTargetLabels:
- service
- servicename

prometheusRule:
name : "{{ tpl $.Values.raa.serviceName $ }}-alerts"
alerts:
health:
for: "30m"
labels:
service: "{{ $.Values.raa.namespace }}"
component: "{{ $.Chart.Name }}"
release: "{{ tpl $.Values.raa.serviceName $ }}"
severity: "severe"
annotations:
summary: RAA is unhealthy for more than hour
description: |-
Pod {{ "{{ $labels.pod }}" }} reports about unhealthy RAA for more than hour already.
error:
labels:
service: "{{ $.Values.raa.namespace }}"
component: "{{ $.Chart.Name }}"
release: "{{ tpl $.Values.raa.serviceName $ }}"
severity: "warning"
annotations:
summary: RAA error detected
description: |-
Pod {{ "{{ $labels.pod }}" }} reports about new RAA errors detected.
longAggregation:
thresholdSec: 300
labels:
service: "{{ $.Values.raa.namespace }}"
component: "{{ $.Chart.Name }}"
release: "{{ tpl $.Values.raa.serviceName $ }}"
severity: "warning"
annotations:
summary: Aggregation query is pretty slow
description: |-
Pod {{ "{{ $labels.pod }}" }} reports that query {{ "{{ $labels.hierarchy }}-{{ $labels.level }}--{{ $labels.mediaType }}" }}
was running more than {{ $.Values.raa.prometheusRule.alerts.longAggregation.thresholdSec }} seconds.
</source>

===Monitoring summary===
|ConfigureMetrics=The metrics that are exposed by these services are available by default. No further configuration is required in order to define or expose these metrics. You cannot define your own custom metrics.
|AlertsDefined=Yes
|ExtConfigAlertsBoilerplate=Yes
|ConfigureAlerts=Set the severity levels and some thresholds for alerts by specifying various '''raa.prometheusRule.alerts.*''' parameters in the '''values.yaml''' file. See {{Link-SomewhereInThisVersion|manual=GCXIPEGuide|topic=RAAMetrics|anchor=Alerts|display text=RAA alerts}} for details.
|Logging=For more information, see {{Link-SomewhereInThisVersion|manual=GCXIPEGuide|topic=Logging}}
}}

@@ Line 11: / Line 11: @@
 |EnableMonitoring=Monitoring is not enabled by default for GCXI or RAA.
 ===Enable GCXI monitoring===
-GCXI Helm Chart ships standard ServiceMonitor object for integration with Prometheus Operator, in [https://git.scm.genesys.com/Genesys/gcxi-docker/blob/dev/k8s/helm/gcxi/templates/gcxi-service-monitor.yaml gcxi-service-monitor.yaml].
+The GCXI Helm Chart provides a standard ServiceMonitor object for integration with Prometheus Operator, in [https://git.scm.genesys.com/Genesys/gcxi-docker/blob/dev/k8s/helm/gcxi/templates/gcxi-service-monitor.yaml gcxi-service-monitor.yaml].
-Related Helm Chart values:
+The following Helm Chart values indicate whether to deploy ServiceMonitor object for integration with Prometheus Operator:
 <source lang="bash">gcxi.deployment.deployServiceMonitor = false|true
-</source>whether to deploy ServiceMonitor object for integration with Prometheus Operator
+</source><br />
 <source lang="bash">gcxi.ports.worker.metrics = 8180 (default)
 container port for metrics endpoint</source>
 ===Enable RAA monitoring===
-RAA exposes metrics with the help of a sidecar container.
+RAA makes metrics available with the help of a sidecar container.
-#To launch the sidecar container, you specify the '''raa.statefulset.containers.monitor''' element (including ports for metrics and/or health endpoints) in '''values.yaml'''. The value of '''raa.env.STAT_SCRAPE_INTERVAL''' must exceed the scrape interval, in seconds, configured in Prometheus for the metrics port.
+#To start the sidecar container, you specify the '''raa.statefulset.containers.monitor''' element, including ports for metrics and/or health endpoints, in the '''values.yaml''' file. Ensure that the value of '''raa.env.STAT_SCRAPE_INTERVAL''' (in seconds) exceeds the scrape interval configured in Prometheus for the metrics port.
 #:For example:
 #:<source lang="bash">
@@ Line 44: / Line 44: @@
 	      ...
 </source>
-#If Prometheus Operator is configured in your Kubernetes environment, then you can configure Prometheus using the parameters '''PodMonitor''' and '''PrometheusRule''' in your '''values.yaml''' file. When configuring '''interval''' and '''scrapeTimeout''' for the health port, take into account that health check usually takes around a minute.
+#If Prometheus Operator is configured in your Kubernetes environment, then you can configure Prometheus using the parameters '''PodMonitor''' and '''PrometheusRule''' in your '''values.yaml''' file. When configuring '''interval''' and '''scrapeTimeout''' for the health port, keep in mind that health check usually takes around a minute.
 #:For example, set the following values:
 #:<source lang="bash">
@@ Line 102: / Line 102: @@
 ===Monitoring summary===
-|ConfigureMetrics=The metrics that are exposed by these services are available by default. No further configuration is required in order to define or expose these metrics. You cannot define your own custom metrics.
+|ConfigureMetrics=The metrics that are made available by these services are available by default. No further configuration is required to define or make available these metrics. You cannot define your own custom metrics.
 |AlertsDefined=Yes
 |ExtConfigAlertsBoilerplate=Yes

PEC-REP/Current/GCXIPEGuide/Observability - Revision history

WikiSysop at 20:15, April 7, 2022

Tgilchri: Published