Monitoring temporal using prometheus-operator
The operator provides support for prometheus-operator to monitor your Temporal cluster. When metrics exposition is enabled with prometheus and serviceMonitor, the operator create a ServiceMonitor for each temporal components (frontend, history, matching & worker).
Enabling metrics exposition to prometheus using prometheus-operator
First of all, your need the prometheus operator running in your cluster, if not you can install a development version on your cluster:
helm repo add prometheus-community https://prometheus-community.github.io/helm-charts
helm repo update
helm install prometheus-operator prometheus-community/kube-prometheus-stack --set prometheus.prometheusSpec.serviceMonitorSelectorNilUsesHelmValues=false --set=prometheusOperator.namespaces.additional={demo,default,}
Then create your temporal cluster and enable the ServiceMonitor
creation:
apiVersion: temporal.io/v1beta1
kind: TemporalCluster
metadata:
name: prod
namespace: demo
spec:
version: 1.24.3
numHistoryShards: 1
# [...]
metrics:
enabled: true
prometheus:
listenPort: 9090
scrapeConfig:
serviceMonitor:
enabled: true
That's all, Prometheus is now scraping temporal components. You can find some grafana dashboards in the official temporal dashboard repository.
Relabeling metrics
For some use cases, you may want to add relabelConfig to the created ServiceMonitors
.
You can use the spec.metrics.prometheus.scrapeConfig.serviceMonitor.metricRelabelings
field.
For instance, to prefix all metrics:
apiVersion: temporal.io/v1beta1
kind: TemporalCluster
metadata:
name: prod
namespace: demo
spec:
# [...]
metrics:
enabled: true
prometheus:
listenPort: 9090
scrapeConfig:
serviceMonitor:
enabled: true
metricRelabelings:
- sourceLabels: [__name__]
targetLabel: __name__
replacement: temporal_$1
To see all the features provided by this field check the monitoring.coreos.com/v1.RelabelConfig
API reference on prometheus-operator website.