Configure delegate metrics
This topic explains how to configure the Prometheus monitoring tool for metrics collection and the Grafana analytics tool for metrics display.
Harness captures delegate agent metrics for delegates with an immutable image type. This process requires a delegate an immutable image. For more information, go to Delegate image types.
The delegate is instrumented for the collection of the following delegate agent metrics.
Metric name | Description |
---|---|
io_harness_custom_metric_task_execution_time | The time it takes to complete a task (in seconds). |
io_harness_custom_metric_tasks_currently_executing | The number of tasks underway. |
io_harness_custom_metric_task_timeout | The number of tasks that time out before completion. |
io_harness_custom_metric_task_completed | The number of tasks completed. |
io_harness_custom_metric_task_failed | The number of failed tasks. |
io_harness_custom_metric_task_rejected * | The number of tasks rejected because of a high load on the delegate. |
io_harness_custom_metric_delegate_connected | Indicates whether the delegate is connected. Values are 0 (disconnected) and 1 (connected). |
io_harness_custom_metric_resource_consumption_above_threshold * | Delegate cpu/memory is above a threshold (defaults to 80%). Provide DELEGATE_RESOURCE_THRESHOLD as the env variable in the delegate YAML to configure the threshold. For more information, go to Configure delegate resource threshold. |
Metrics notated with * above only visible if you start your delegate with DYNAMIC_REQUEST_HANDLING
set to true
in your delegate YAML. Go to Configure delegate resource threshold for more information.
Also note that the above metrics are available only if your delegate version is later than 23.05.79311.
This topic includes example YAML files you can use to create application manifests for your Prometheus and Grafana configurations.
Apply the prometheus.yml file
The configuration of Prometheus requires the installation of a Prometheus workload and service in your Kubernetes cluster. Use the following example configuration file to install the harness-delegate-prometheus-deployment
workload and a service named harness-delegate-prometheus-service
. The configuration includes a load balancer with an IP address you can use to access the Prometheus UI.
Example prometheus.yml file
apiVersion: v1
kind: Namespace
metadata:
name: harness-delegate-ng
---
apiVersion: v1
kind: ConfigMap
metadata:
name: prometheus-delegate-conf
labels:
name: prometheus-delegate-conf
namespace: harness-delegate-ng
data:
CPU: "1"
MEMORY: "2048"
POD_MEMORY: "3072"
prometheus.yml: |-
global:
scrape_interval: 10s
evaluation_interval: 10s
scrape_configs:
- job_name: 'kubernetes-apiservers'
kubernetes_sd_configs:
- role: endpoints
scheme: http
metrics_path: '/api/metrics'
relabel_configs:
- source_labels: [__meta_kubernetes_namespace, __meta_kubernetes_service_name, __meta_kubernetes_endpoint_port_name]
action: keep
regex: default;kubernetes;https
- job_name: 'kubernetes-pods'
kubernetes_sd_configs:
- role: pod
relabel_configs:
- source_labels: [__meta_kubernetes_pod_annotation_prometheus_io_scrape]
action: keep
regex: true
- source_labels: [__meta_kubernetes_pod_annotation_prometheus_io_path]
action: replace
target_label: __metrics_path__
regex: (.+)
- source_labels: [__address__, __meta_kubernetes_pod_annotation_prometheus_io_port]
action: replace
regex: ([^:]+)(?::\d+)?;(\d+)
replacement: $1:$2
target_label: __address__
- action: labelmap
regex: __meta_kubernetes_pod_label_(.+)
- source_labels: [__meta_kubernetes_namespace]
action: replace
target_label: kubernetes_namespace
- source_labels: [__meta_kubernetes_pod_name]
action: replace
target_label: kubernetes_pod_name
- job_name: 'kubernetes-service-endpoints'
kubernetes_sd_configs:
- role: endpoints
relabel_configs:
- source_labels: [__meta_kubernetes_service_annotation_prometheus_io_scrape]
action: keep
regex: true
- source_labels: [__meta_kubernetes_service_annotation_prometheus_io_scheme]
action: replace
target_label: __scheme__
regex: (https?)
- source_labels: [__meta_kubernetes_service_annotation_prometheus_io_path]
action: replace
target_label: __metrics_path__
regex: (.+)
- source_labels: [__address__, __meta_kubernetes_service_annotation_prometheus_io_port]
action: replace
target_label: __address__
regex: ([^:]+)(?::\d+)?;(\d+)
replacement: $1:$2
- action: labelmap
regex: __meta_kubernetes_service_label_(.+)
- source_labels: [__meta_kubernetes_namespace]
action: replace
target_label: kubernetes_namespace
- source_labels: [__meta_kubernetes_service_name]
action: replace
target_label: kubernetes_name
---
apiVersion: v1
kind: Service
metadata:
name: harness-delegate-prometheus-service
namespace: harness-delegate-ng
spec:
selector:
app: prometheus-delegate
type: LoadBalancer
ports:
- port: 8084
targetPort: 9090
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: harness-delegate-prometheus-deployment
namespace: harness-delegate-ng
spec:
replicas: 1
template:
metadata:
labels:
app: prometheus-delegate
spec:
containers:
- name: prometheus
image: prom/prometheus:v2.6.0
args:
- "--config.file=/etc/prometheus/prometheus.yml"
- "--storage.tsdb.path=/prometheus/"
ports:
- containerPort: 9090
volumeMounts:
- name: prometheus-config-volume
mountPath: /etc/prometheus/
- name: prometheus-storage-volume
mountPath: /prometheus/
volumes:
- name: prometheus-config-volume
configMap:
defaultMode: 420
name: prometheus-delegate-conf
- name: prometheus-storage-volume
emptyDir: {}
selector:
matchLabels:
app: prometheus-delegate
Use the following command to deploy the configuration file.
kubectl apply -f prometheus.yml
Set up Grafana
To set up Grafana, use the following example grafana.yml
file.
Example grafana.yml file
apiVersion: v1
kind: ConfigMap
metadata:
name: grafana-datasources
namespace: harness-delegate-ng
data:
prometheus.yaml: |-
{
"apiVersion": 1,
"datasources": [
{
"access":"proxy",
"editable": true,
"name": "prometheus",
"orgId": 1,
"type": "prometheus",
"url": "http://harness-delegate-prometheus-service:8084",
"version": 1
}
]
}
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: grafana
namespace: harness-delegate-ng
spec:
replicas: 1
selector:
matchLabels:
app: grafana
template:
metadata:
name: grafana
labels:
app: grafana
spec:
containers:
- name: grafana
image: grafana/grafana:latest
ports:
- name: grafana
containerPort: 3000
resources:
limits:
memory: "1Gi"
cpu: "1000m"
requests:
memory: 500M
cpu: "500m"
volumeMounts:
- mountPath: /var/lib/grafana
name: grafana-storage
- mountPath: /etc/grafana/provisioning/datasources
name: grafana-datasources
readOnly: false
volumes:
- name: grafana-storage
emptyDir: {}
- name: grafana-datasources
configMap:
defaultMode: 420
name: grafana-datasources
---
apiVersion: v1
kind: Service
metadata:
name: grafana
namespace: harness-delegate-ng
annotations:
prometheus.io/scrape: 'true'
prometheus.io/port: '3000'
spec:
selector:
app: grafana
type: LoadBalancer
ports:
- port: 3000
targetPort: 3000
Copy the
grafana.yml
file.If you're not using the default
harness-delegate-ng
namespace, replace it with the namespace into which you deployed your delegate.Use the following command to apply the Grafana configuration file to your deployment:
kubectl apply -f grafana.yml
noteThis manifest also creates a load balancer and service in your Kubernetes cluster.
Select the exposed URL to access Grafana.
Configure delegate resource threshold
You can set the delegate to reject new tasks if x% of memory is being consumed. You can then spin up new delegates when resources are above the threshold.
The io_harness_custom_metric_resource_consumption_above_threshold
metric is only visible if you start your delegate with DYNAMIC_REQUEST_HANDLING set to true
in your delegate YAML.
To configure the delegate resource threshold, make the following changes to the delegate YAML file:
Set the
JAVA_OPTS
env variable.env:
- name: JAVA_OPTS
value: "-Xmx1536M"Set the
DYNAMIC_REQUEST_HANDLING
env variable totrue
to enable dynamic task rejection.env:
- name: DYNAMIC_REQUEST_HANDLING
value: "true"Set the
RESOURCE_USAGE_THRESHOLD
env variable to the cpu/memory threshold. When the threshold is exceeded, the delegate rejects new tasks.env:
- name: RESOURCE_USAGE_THRESHOLD
value: "80"noteThis step is optional if you want to use the default value of 80%.