README
¶
tinySQL Catalog and Scheduler Demo
This example registers tables, views, functions, and scheduled SQL jobs in the tinySQL catalog. It has a short terminal mode and a browser dashboard for a running scheduler.
Run
# terminal demo; finishes after about six seconds
go run ./cmd/catalog_demo
# browser dashboard
go run ./cmd/catalog_demo -web -addr 127.0.0.1:8089
# http://localhost:8089
| Flag | Default | Purpose |
|---|---|---|
| -web | false | Keep the scheduler running and serve the dashboard |
| -addr | 127.0.0.1:8089 | Dashboard listen address |
The terminal demo runs refresh_event_stats every two seconds and runs integrity_check once. In web mode, refresh_event_stats runs every 15 seconds and integrity_check runs once after one minute. The dashboard can manually run only these registered jobs; it accepts no arbitrary SQL.
What it demonstrates
- catalog registration and inspection for tables, views, functions, and jobs;
- interval and one-time job schedules;
- a JobExecutor that parses and executes catalog SQL;
- scheduler start, stop, status, and recent execution history.
The web mode starts with a fresh in-memory dataset. Tables, job history, and scheduler state disappear on restart.
HTTP API
| Method | Path | Purpose |
|---|---|---|
| GET | /healthz | Liveness check |
| GET | /api/state | Catalog, job status, and recent runs |
| POST | /api/jobs/{name}/run | Run a registered job |
curl -X POST http://127.0.0.1:8089/api/jobs/refresh_event_stats/run
The dashboard has no login or TLS. Keep it on loopback, or add authentication and HTTPS before making it reachable over a network.
Documentation
¶
Overview ¶
catalog_demo – tinySQL Catalog & Scheduler demonstration.
This tool shows how to use tinySQL's built-in catalog (table/view/function registry) and job scheduler to run periodic SQL maintenance tasks.
The scheduler executor is backed by the real tinySQL engine, so every scheduled job actually runs its SQL statement and prints the result.
Usage:
catalog_demo
Source Files