---
title: "Infrastructure"
description: "Monitor containers in Middleware to track resource usage, performance, and infrastructure health effectively."
section: "Infrastructure"
breadcrumb: "Infrastructure > Container Monitoring"
canonical: "https://docs.middleware.io/features/infrastructure"
---
# Infrastructure

Infrastructure monitoring starts with the Host Agent. Install it once per machine and it reports CPU, memory, disk and network for the host, every process the OS is running, and every container on that box, with no extra config per service. On Kubernetes the same agent runs as a DaemonSet and walks the cluster for you: nodes, namespaces, workloads and pods show up already linked to the containers underneath them. On ECS it rides along as a sidecar in the task definition, on EC2 or Fargate, and pulls task and container metrics from the ECS metadata endpoint. The hierarchy comes from where the agent is running, not from you tagging hosts to clusters by hand.

## Features

- [Host](/features/infrastructure/hosts) — CPU, memory, disk and network for every machine
- [Process](/features/infrastructure/processes) — See which processes are using the host's CPU and memory
- [Kubernetes](/features/infrastructure/kubernetes) — Clusters, nodes, workloads and pods from a DaemonSet install
- [ECS](/features/infrastructure/ecs) — Tasks and containers on EC2 or Fargate via a sidecar agent

## Related Features

- [Logs Management](/features/logs) — Search, filter and turn log patterns into metric
- [APM](/features/apm) — Traces, services, profiles and endpoints
- [Dashboards](/features/dashboards) — Visualise telemetry data
- [Alerts](/features/alerts) — Create alert rules and service level objectives