# UI Dashboard

Logbus includes a built-in web dashboard for observing a running pipeline and the host it is running on. It is disabled by default and adds no overhead when not in use.

## Enabling the Dashboard

Pass `--ui` to start the dashboard on the default port (9100):

```sh
logbus --ui pipeline.yml
```

To use a different port, pass `--ui-port` (this also enables the dashboard):

```sh
logbus --ui-port 9200 pipeline.yml
```

Once running, open `http://localhost:9100` (or your chosen port) in a browser.

## Pipeline Observability

### Per-stage Stats

The dashboard shows live counters for every stage, updated every second:

| Metric | Description |
|:---|:---|
| `rx-events` | Events received by the stage |
| `rx-bytes` | Bytes received |
| `tx-events` | Events emitted by the stage |
| `tx-bytes` | Bytes emitted |
| `errors` | Error events produced |

### Event Samples

For each stage you can inspect the last 10 events that passed through it. This is useful for verifying that parsing, enrichment, or filtering is working as expected without needing to add a separate debug sink.

### Recent Errors

The dashboard keeps a rolling log of the last 50 error events across all stages, each annotated with the stage that produced it. See the [ERRORS channel](/features/errors-channel) for details on how errors are produced.

## Host Observability

The dashboard also surfaces metrics about the host running the pipeline:

- **Host Info** — hostname, OS name and version, uptime
- **Load Average** — 1, 5, and 15-minute averages
- **CPU** — overall usage (%) and physical core count
- **Memory** — used and total
- **Network** — per-interface bytes received/transmitted (since last refresh and total)
- **Process RSS** — memory currently used by the logbus process itself