# Grafatui Examples
This directory contains example Grafana dashboards and a demo environment for testing grafatui.
## Quick Demo
Want to try grafatui instantly? Use the pre-configured demo environment:
```bash
cd demo
docker-compose up -d && sleep 5 && cd ../.. && cargo run -- --grafana-json examples/dashboards/prometheus_demo.json --prometheus-url http://localhost:19090
```
See [`demo/README.md`](demo/README.md) for details.
## External Annotations
[`annotations.jsonl`](annotations.jsonl) is a read-only external JSONL event
source for graph and timeseries panels. Its `Deployed v2.4` event uses
`panel_titles` to target the bundled `HTTP Request Rate by Status Code` panel;
the remaining events omit `panel_titles` and therefore remain dashboard-wide.
Run it with a dashboard:
```bash
cargo run -- --grafana-json examples/dashboards/prometheus_demo.json --annotations-file examples/annotations.jsonl
```
See the [external annotations guide](../docs/annotations.md) for targeting,
tag filtering, inspection, reload behavior, and limitations.
### Git Command Provider
`git_annotation_provider.rs` turns commits in Grafatui's requested time window
into external JSONL annotations. Build it from the repository root:
```bash
cargo build --example git_annotation_provider
```
Then configure the compiled example as the one annotation source:
```toml
[annotations_command]
program = "./target/debug/examples/git_annotation_provider"
args = ["."]
timeout = "10s"
```
The optional first argument is the Git repository path and defaults to `.`.
## Dashboards
### `prometheus_demo.json`
**Recommended for first-time users!** A comprehensive dashboard designed for the included demo environment.
Shows all 6 visualization types with real metrics from Prometheus monitoring itself:
- Graph, Gauge, Stat, Bar Gauge, Table, Heatmap
- Works immediately with `demo/docker-compose.yml`
### `all_visualizations.json`
Demonstrates all supported panel types in a single dashboard:
- **Graph**: Line chart showing CPU usage over time
- **Timeseries styles**: Bars with hidden axis, area fill, and points-only rendering
- **Gauge**: Progress bar for memory usage
- **Stat**: Big value display with sparkline for uptime
- **Bar Gauge**: Vertical bars comparing request rates
- **Table**: Tabular view of series
- **Heatmap**: Color-coded time-series intensity
### `instant_queries.json`
Demonstrates explicit `targets[].instant` support and the default instant query
behavior used by Gauge, Bar Gauge, and Table panels.
### `thresholds_demo.json`
Demonstrates threshold rendering, threshold marker styles, and explicit field
min/max bounds.
### Usage
```bash
# Test with local Prometheus (default port 9090)
cargo run -- --grafana-json examples/dashboards/all_visualizations.json
# Or with custom Prometheus URL
cargo run -- --grafana-json examples/dashboards/all_visualizations.json --prometheus-url http://prometheus.example.com:9090
# Override variables
cargo run -- --grafana-json examples/dashboards/all_visualizations.json --var instance=prometheus:9090
```
## Creating Your Own
You can export any Grafana dashboard as JSON and use it with grafatui:
1. In Grafana, go to Dashboard Settings → JSON Model
2. Copy the JSON
3. Save it to a file
4. Run: `grafatui --grafana-json your-dashboard.json`
## Supported Panel Types
- ✅ `graph` / `timeseries` - Line, point, bar, and area-style charts
- ✅ `gauge` - Progress bars
- ✅ `bargauge` - Bar charts
- ✅ `table` - Data tables
- ✅ `stat` - Single value + sparkline
- ✅ `heatmap` - Color intensity grid