1. Documentation
×

Monitoring the Network Observability Operator

Use the OKD web console to monitor alerts related to the Network Observability Operator’s health. This helps you maintain system stability and quickly detect operational issues.

Health dashboards

View the Network Observability Operator health dashboards in the OKD web console to monitor the health status, resource usage, and internal statistics of the operator and its components.

Metrics are located in the Observe → Dashboards page in the OKD web console. You can view metrics about the health of the Network Observability Operator in the following categories:

  • Flows per second

  • Sampling

  • Errors last minute

  • Dropped flows per second

  • Flowlogs-pipeline statistics

  • Flowlogs-pipleine statistics views

  • eBPF agent statistics views

  • Operator statistics

  • Resource usage

Health alerts

Understand the health alerts generated by the Network Observability Operator, which trigger banners when conditions like Loki ingestion errors, zero flow ingestion, or dropped eBPF flows occur.

A health alert banner that directs you to the dashboard can appear on the Network Traffic and Home pages if an alert is triggered. Alerts are generated in the following cases:

  • The NetObservLokiError alert occurs if the flowlogs-pipeline workload is dropping flows because of Loki errors, such as if the Loki ingestion rate limit has been reached.

  • The NetObservNoFlows alert occurs if no flows are ingested for a certain amount of time.

  • The NetObservFlowsDropped alert occurs if the Network Observability eBPF agent hashmap table is full, and the eBPF agent processes flows with degraded performance, or when the capacity limiter is triggered.

Viewing health information

View the Netobserv/Health dashboard within the OKD web console to monitor the health status and resource usage of the Network Observability Operator and its components.

Prerequisites

  • You have the Network Observability Operator installed.

  • You have access to the cluster as a user with the cluster-admin role or with view permissions for all projects.

Procedure

  1. From the Administrator perspective in the web console, navigate to Observe → Dashboards.

  2. From the Dashboards dropdown, select Netobserv/Health.

  3. View the metrics about the health of the Operator that are displayed on the page.

Disabling health alerts

Disable specific health alerts, such as NetObservLokiError or NetObservNoFlows, by editing the FlowCollector resource and using the spec.processor.metrics.disableAlerts specification.

Procedure

  1. In the web console, navigate to Ecosystem → Installed Operators.

  2. Under the Provided APIs heading for the NetObserv Operator, select Flow Collector.

  3. Select cluster then select the YAML tab.

  4. Add spec.processor.metrics.disableAlerts to disable health alerts, as in the following YAML sample:

    apiVersion: flows.netobserv.io/v1beta2
kind: FlowCollector
metadata:
  name: cluster
spec:
  processor:
    metrics:
      disableAlerts: [NetObservLokiError, NetObservNoFlows] (1)
    1 You can specify one or a list with both types of alerts to disable.

Creating Loki rate limit alerts for the NetObserv dashboard

Create a custom AlertingRule resource based on Loki metrics to monitor for and trigger alerts when the Loki ingestion rate limits are reached, indicated by HTTP 429 errors.

You can create custom alerting rules for the Netobserv dashboard metrics to trigger alerts when Loki rate limits have been reached.

Prerequisites

  • You have access to the cluster as a user with the cluster-admin role or with view permissions for all projects.

  • You have the Network Observability Operator installed.

Procedure

  1. Create a YAML file by clicking the import icon, +.

  2. Add an alerting rule configuration to the YAML file. In the YAML sample that follows, an alert is created for when Loki rate limits have been reached:

    apiVersion: monitoring.openshift.io/v1
kind: AlertingRule
metadata:
  name: loki-alerts
  namespace: openshift-monitoring
spec:
  groups:
  - name: LokiRateLimitAlerts
    rules:
    - alert: LokiTenantRateLimit
      annotations:
        message: |-
          {{ $labels.job }} {{ $labels.route }} is experiencing 429 errors.
        summary: "At any number of requests are responded with the rate limit error code."
      expr: sum(irate(loki_request_duration_seconds_count{status_code="429"}[1m])) by (job, namespace, route) / sum(irate(loki_request_duration_seconds_count[1m])) by (job, namespace, route) * 100 > 0
      for: 10s
      labels:
        severity: warning

  3. Click Create to apply the configuration file to the cluster.

Using the eBPF agent alert

Resolve the NetObservAgentFlowsDropped alert, which occurs when the eBPF agent hashmap is full, by increasing the spec.agent.ebpf.cacheMaxFlows value in the FlowCollector custom resource.

An alert, NetObservAgentFlowsDropped, is also triggered when the capacity limiter is triggered. If you see this alert, consider increasing the cacheMaxFlows in the FlowCollector, as shown in the following example.

Increasing the cacheMaxFlows might increase the memory usage of the eBPF agent.
Procedure

  1. In the web console, navigate to Ecosystem → Installed Operators.

  2. Under the Provided APIs heading for the Network Observability Operator, select Flow Collector.

  3. Select cluster, and then select the YAML tab.

  4. Increase the spec.agent.ebpf.cacheMaxFlows value, as shown in the following YAML sample:

apiVersion: flows.netobserv.io/v1beta2
kind: FlowCollector
metadata:
  name: cluster
spec:
  namespace: netobserv
  deploymentModel: Direct
  agent:
    type: eBPF
    ebpf:
      cacheMaxFlows: 200000 (1)
1 Increase the cacheMaxFlows value from its value at the time of the NetObservAgentFlowsDropped alert.
Additional resources