⚡

Rust + HTMX • ScyllaDB & ClickHouse HTAP

ScyllaDB (OLTP)

ClickHouse (OLAP)

Navigation

Dashboard Collections & Schema Data Explorer Auth & Users ScyllaDB Console Storage & Files Realtime CDC AI Vector Search Webhooks & Triggers Analytics & SQL

HTAP Topology

Single write path auto-replicates row state to ScyllaDB for consistency and streams CDC to ClickHouse for real-time analytics.

System Dashboard

Real-time telemetry and dual-engine performance

New Collection

OLTP Records

Stored in ScyllaDB

OLAP Events Ingested

Aggregated by ClickHouse

Avg OLTP Latency

1.08 ms

Sub-ms Point Reads/Writes

Avg OLAP Latency

3.18 ms

Vectorized Columnar Execution

Event Throughput Timeline

Real-time request distribution streamed to ClickHouse

OLAP Time-Bucket

Engine Capability Split

ScyllaDB Tasks OLTP

• Primary Key point lookups
• Fast record inserts & updates
• User credentials & auth tokens
• KeySpace partition indexing

ClickHouse Tasks OLAP

• Real-time log aggregation
• Vectorized multi-column filters
• ReplacingMergeTree replica queries
• Time-series analytics & metrics

Database REST API Reference & Endpoints

Direct HTTP endpoints to access ScyllaDB (OLTP) and ClickHouse (OLAP)

API Base Host: http://localhost:3000/api/v1

Master Service Key: bk_admin_service_role_secret_key_100

HTTP Method	API Endpoint Path	Engine	Action
GET	/api/v1/collections	ScyllaDB	Test API →
POST	/api/v1/collections/:col/records	Dual (ScyllaDB+CDC)	Insert Record
GET	/api/v1/collections/:col/records/:id	ScyllaDB (<1ms)	Point Lookup
GET	/api/v1/collections/:col/records	ScyllaDB / ClickHouse	Query Records
DELETE	/api/v1/collections/:col/records/:id	ScyllaDB CDC	Delete Record
POST	/api/v1/analytics/sql	ClickHouse (OLAP)	Vector SQL Query

JavaScript Fetch Example:

const response = await fetch('http://localhost:3000/api/v1/collections/posts/records', {
  method: 'POST',
  headers: {
    'Content-Type': 'application/json',
    'x-api-key': 'bk_admin_service_role_secret_key_100'
  },
  body: JSON.stringify({
    title: 'Hello BaaS Lemnidev Database',
    views: 120
  })
});
const data = await response.json();
console.log(data);