Infrastructure Monitoring

Overview

Osuite monitors the health of your hosts and Kubernetes clusters using the OpenTelemetry Collector — no proprietary agents and no vendor-specific exporters. Infrastructure metrics arrive over OTLP like every other signal, which means they are correlated with your application data: from a degraded service in APM you can navigate straight to the host or pod metrics underneath it, and the Investigation Agent can pin whether an incident is the application’s fault or the machine’s.

Because ingestion is OpenTelemetry-native, anything that can emit OTel metrics — hosts, Kubernetes, cloud services (AWS, GCP, Azure), databases, message queues — shows up in Osuite without a custom integration.

What you can monitor

Host monitoring

Track CPU, memory, disk, network, and load across every host in real time, with pre-built host dashboards ready the moment metrics start flowing. Use the Collector’s hostmetrics receiver:

receivers:
  hostmetrics:
    collection_interval: 30s
    scrapers:
      cpu:
      memory:
      disk:
      filesystem:
      network:
      load:

processors:
  batch:
  resourcedetection:

exporters:
  otlp/osuite:
    endpoint: ingest.<region>.osuite.io:443
    tls:
      insecure: false
    headers:
      x-osuite-ingest-token: <your ingest token>

service:
  pipelines:
    metrics:
      receivers: [hostmetrics]
      processors: [batch, resourcedetection]
      exporters: [otlp/osuite]

The resourcedetection processor attaches host and cloud identity (hostname, instance ID, region) so each metric is attributable to the machine that produced it.

Kubernetes monitoring

Get full visibility into cluster, node, namespace, and pod health, with pre-built Kubernetes dashboards covering the metrics platform teams care about. The standard path is the OpenTelemetry Collector deployed in your cluster:

• The kubeletstats receiver collects node, pod, and container resource metrics.
• The k8s_cluster receiver collects cluster-level state (deployments, node conditions, pod phases).
• The k8sattributes processor enriches every metric, log, and trace with k8s.namespace.name, k8s.pod.name, k8s.deployment.name, and other identity, so infrastructure metrics line up with the application telemetry from the same pods.

Deploy it with the OpenTelemetry Collector Helm chart, point the otlp/osuite exporter at your ingest endpoint, and the pre-built dashboards populate automatically.

Ingest any infrastructure metric

Anything that speaks OpenTelemetry can be visualized in Osuite. Add a receiver for the source — a cloud service, a database, a queue — to your Collector config and route it through the same otlp/osuite exporter. Extend the pre-built dashboards or build new ones with the Custom Dashboard builder.

Correlated with your applications

Infrastructure metrics aren’t a separate silo. They share resource attributes with your APM, logs, and traces, so during an incident you can move seamlessly between a degraded service and the host or pod metrics beneath it — and the Investigation Agent correlates the two automatically when it builds a timeline.

Next steps

• APM & Distributed Tracing — correlate infrastructure health with service performance
• Dashboards — customize the pre-built host and Kubernetes dashboards, or build your own
• Alerts — alert on resource thresholds before utilization becomes an incident