> ## Documentation Index
> Fetch the complete documentation index at: https://www.towbar.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Monitoring, health, and capacity

> Read control-plane checks, runtime health, resource usage, and operation history.

Towbar exposes several kinds of health. Read the relevant check and its timestamp before deciding whether an installation or workload is ready.

## Navigate monitoring

The workspace **Monitor** section contains **Incidents** and **Vulnerabilities**. On Incidents, choose **Active** or **Resolved**, then optionally filter by severity. On mobile, open **Page menu** for these controls.

## Performance history

Open a server, app, or resource and select **Monitor → Performance**. Use the time range and chart controls to inspect CPU, memory, disk, and network measurements. Workload views also offer environment and container selection. The URL retains the chart filters, so a shared link opens the same view.

Deployment and restart markers connect changes in usage to operational events. Drag across a chart to zoom, or choose a custom time range. Missing measurements appear as gaps. See [Scout Agent](/docs/scout) for collection and retention, and [Deployment comparisons](/docs/deployment-comparisons) for comparing releases over equal windows.

## Alerts, incidents, and vulnerabilities

Use workspace **Incidents** to review active or resolved conditions across every server, app, and resource. Filter the list by severity. Configure and inspect alert rules from the relevant entity page; [Scout Alerts](/docs/scout-alerts) covers rules and notification destinations. **Vulnerabilities** provides a separate view of known package findings from app image scans; see [Vulnerability scanning](/docs/vulnerability-scanning).

## System health

Open **Manage → System health** and choose **Run checks**.

| Section       | Checks                                     | What to investigate                                                                 |
| ------------- | ------------------------------------------ | ----------------------------------------------------------------------------------- |
| Control plane | API/database, Temporal, worker/maintenance | API reachability, workflow admission, recent worker activity, and version alignment |

<div className="towbar-doc-screenshot">
  <div className="towbar-product-light">
    <img src="https://mintcdn.com/avgeek/-D5mUM-1hnhKZAmI/assets/release-v2/system-health-light.jpg?fit=max&auto=format&n=-D5mUM-1hnhKZAmI&q=85&s=4ad40fdd85cacfb37574d38ae24aa8c6" alt="System health reports only control-plane checks and their latest observation time." width="1280" height="720" loading="lazy" data-path="assets/release-v2/system-health-light.jpg" />
  </div>

  <div className="towbar-product-dark">
    <img src="https://mintcdn.com/avgeek/-D5mUM-1hnhKZAmI/assets/release-v2/system-health-dark.jpg?fit=max&auto=format&n=-D5mUM-1hnhKZAmI&q=85&s=145602229f684589925b7d0d94136bf5" alt="System health reports only control-plane checks and their latest observation time." width="1280" height="720" loading="lazy" data-path="assets/release-v2/system-health-dark.jpg" />
  </div>

  <p>
    System health focuses on Towbar's control plane; provider availability lives
    under Integrations.
  </p>
</div>

## Server and workload health

The server's Checks view records connectivity and runtime observations. Its Apps and Resources sidebar items separate the workloads assigned to the host. Health indicators and tooltips identify healthy, unhealthy, starting, and unavailable observations.

A deployment can succeed and later become unhealthy. Check the latest runtime state and logs instead of relying only on deployment history.

## Allocated CPU and memory

Allocation is the limit declared for a workload. Usage is the latest observed consumption relative to that allocation. CPU percentages account for the number of allocated CPUs; memory percentages use the configured memory limit.

Server capacity columns describe the host's maximum capacity. They do not represent an app's allocation. Missing measurements display as unavailable rather than zero. Review the observation timestamp if the server cannot be reached.

<div className="towbar-doc-screenshot">
  <div className="towbar-product-light">
    <img src="https://mintcdn.com/avgeek/7IKQY5hBmeQaVMw0/assets/release-v2/server-workloads-light.jpg?fit=max&auto=format&n=7IKQY5hBmeQaVMw0&q=85&s=1949e5b094a6adb0dfadbcb8e253611a" alt="A server’s Apps page compares observed usage with allocated CPU and memory." width="1280" height="720" loading="lazy" data-path="assets/release-v2/server-workloads-light.jpg" />
  </div>

  <div className="towbar-product-dark">
    <img src="https://mintcdn.com/avgeek/7IKQY5hBmeQaVMw0/assets/release-v2/server-workloads-dark.jpg?fit=max&auto=format&n=7IKQY5hBmeQaVMw0&q=85&s=aeab68ae2b6a63c0ed53591968a124e6" alt="A server’s Apps page compares observed usage with allocated CPU and memory." width="1280" height="720" loading="lazy" data-path="assets/release-v2/server-workloads-dark.jpg" />
  </div>

  <p>
    A server’s Apps page compares observed usage with allocated CPU and memory.
  </p>
</div>

## Follow an operation

Deployment, preparation, backup, and restore histories show stage progress and final results. Durations update while an operation is running. Date columns show relative time first, with the local date and time below it as **15:26, 10 Sept 2026**. Hover or focus the timestamp to see its timezone.

For runtime problems, inspect the workload's logs and latest health check. For control-plane problems, start with:

```bash theme={"system"}
docker compose ps
docker compose logs --tail 200 api worker temporal
```

Configure [notifications](/docs/integrations/notifications) for events that need attention. Use [Troubleshooting](/docs/troubleshooting) to narrow down common symptoms.

## Scout Agent monitoring agent

[Scout Agent](/docs/scout) adds continuous performance history for servers, apps, and resources. Install it from **Server → Settings → Scout Agent** to receive updates every 30 seconds and keep up to 60 days of history. Read the [Scout Agent guide](/docs/scout) for installation, charts, retention, and troubleshooting.