Skip to content
Start free

Know what runs where, and how hard it is working

One collector reads host metrics on Linux, macOS and Windows, cluster state and pod usage from Kubernetes, and container stats from Docker. Tracix also inventories your AWS, Azure and Google Cloud accounts and marks the resources your collectors already watch.

Start freeBook a demoRead the infrastructure docs
  • Linux · macOS · Windows
  • Kubernetes and OpenShift
  • AWS · Azure · Google Cloud

How it works

From one collector to a map of your estate

Metrics arrive with OpenTelemetry resource names such as host.name and k8s.pod.name, so the same host or pod lines up across metrics, logs and traces.

  1. 01InstallRun the installer on each host, or the Helm chart as a DaemonSet on every Linux node.
  2. 02CollectHost metrics, cluster state, pod and container usage, Kubernetes events and container logs share one pipeline.
  3. 03DiscoverConnect cloud accounts with read-only credentials, and Tracix keeps an inventory of their resources.
  4. 04MatchDiscovered instances and clusters are matched to the hosts and clusters your collectors monitor.
  5. 05WatchMaps and explorers show what is hot, stale or offline, and monitors on any metric tell you when it changes.

Kubernetes

Kubernetes from the cluster down to the container

The Helm chart runs the collector on every Linux node and reads cluster state, pod and container usage, object inventory and events through the OpenTelemetry Kubernetes receivers. Requests, limits and real usage sit side by side, so idle capacity and pods without limits stand out.

  • An explorer for pods, nodes, workloads and services, with metrics, YAML and events for each
  • Resource utilization compares each pod's CPU and memory with its requests and limits
  • With flow capture on, a network view shows traffic between pods with retransmits and round-trip time

Cloud accounts

Every cloud account, inventoried and matched to what you monitor

Connect AWS with a cross-account role, Azure with a Reader service principal and Google Cloud with a Viewer service account. Discovery runs from Tracix with no agent in your accounts, and draws how the resources connect.

  • AWS

    EC2, RDS, load balancers, S3, Lambda, ECS, EKS, CloudFront and Route 53 in every region, plus Cloud Control for more types.

  • Azure

    Virtual machines, networks, SQL, load balancers, storage, Functions, AKS, CDN and DNS.

  • Google Cloud

    Compute Engine, Cloud SQL, load balancing, storage, Cloud Functions, Cloud Run, GKE, CDN and DNS.

  • Cloud metrics

    CloudWatch metrics and logs for AWS and Azure Monitor metrics for Azure, on the same pipeline.

Metrics explorer

Find any metric by its OpenTelemetry name

Every metric in the window shows up as a card. Filter by service, dataset or any attribute, break a chart down by a dimension, and turn a card into a monitor. When you need more, SQL mode runs read-only ClickHouse queries over the metric store.

  • Aggregations fit the metric: percentiles for histograms, rates for counters, averages for gauges
  • Prometheus endpoints feed the same store through the collector's scraper
  • An assistant drafts SQL for you to review and run, when an AI provider is configured

Better together

Infrastructure is where application problems land

Hosts, pods and cloud resources share a workspace with traces, logs and network devices, so you can tell a slow service from a busy node in one place.

Supported platforms

What Tracix collects on each platform

Hosts and clusters report through the Tracix Collector. Cloud accounts are read by agentless discovery and the cloud integrations.

PlatformWhat is collectedHow
LinuxCPU, memory, load, disk, filesystem, network and uptime; processes, paging and NFS when enabledCollector as a systemd service
macOS 12 and laterCPU, memory, load, disk, filesystem, network and uptimeCollector as a launchd daemon
Windows Server 2016 and Windows 10 or laterCPU, memory, disk, filesystem, network and uptime; no process countsCollector as a Windows service
DockerCPU, memory and image for each containerDocker stats over the Docker socket
KubernetesCluster state, pod and container usage against requests and limits, objects, events and node metricsHelm chart, a DaemonSet on Linux nodes
AWSInventory and topology; CloudWatch metrics and logsCross-account role or access key
AzureInventory and topology; Azure Monitor metricsService principal
Google CloudInventory and topology; metrics and logs you export to Pub/SubService account
PrometheusAny /metrics endpointPrometheus scraping

Linux means Ubuntu 20.04+, Debian 11+, RHEL and CentOS 8+ or Amazon Linux 2. Every host platform runs on amd64 and arm64.

FAQ

Questions infrastructure teams ask first

Talk to an engineer
Do I need an agent in my cloud accounts?
No. Cloud discovery runs from Tracix with read-only credentials you provide: a cross-account role or access key for AWS, a service principal with Reader for Azure, and a service account with Viewer for Google Cloud.
Is infrastructure monitoring on the Free plan?
Not on Free. Infrastructure monitoring is part of the Starter, Professional and Enterprise plans, and the pricing page lists what each plan includes.
Does the Helm chart run on Windows nodes?
No. The chart runs the collector as a DaemonSet on Linux nodes only. Windows servers outside Kubernetes use the Windows installer.
Can Tracix read Google Cloud Monitoring directly?
Not today. Tracix inventories Google Cloud projects through their APIs, but metrics and logs arrive only through a Pub/Sub subscription you export them to.

See your first hosts on the map in minutes

Install the collector on a Linux, macOS or Windows host, or add the Helm chart to a cluster, and the host map fills in as metrics arrive.

Start freeRead the install guide