🗓️ 07082025 1042
📎
htap
HTAP = Hybrid Transactional and Analytical Processing
It’s a database architecture that supports both oltp and olap workloads on the same system, often in real time — without needing ETL.
🚀 Why HTAP is a Big Deal
| 🔍 Traditional | 🔥 HTAP |
|---|---|
| OLTP + OLAP are separate | Unified system |
| ETL delays | Real-time sync |
| High infra complexity | Simpler architecture |
| Costly duplication | Single source of truth |
🧩 How Does HTAP Work?
HTAP systems often:
- Store data once, but in formats suited for both workloads
- Use multi-engine architectures (row store + column store)
- Separate compute for isolation (no interference)
- Leverage MVCC, HTAP schedulers, or replication to optimize for both use cases
🛠️ Real-World Examples
| System | HTAP Capabilities |
|---|---|
| TiDB (PingCAP) | OLTP (MySQL-like) + OLAP (TiFlash) |
| SingleStore | Unified row/column engine |
| CockroachDB | OLTP + some analytical support |
| MySQL + HeatWave | HTAP via in-memory analytics |
| ClickHouse | OLAP-focused, but moving toward HTAP with insert speed improvements |
| DorisDB / StarRocks | Emerging HTAP-style performance |
✅ Benefits
- 💡 Real-time analytics on live data
- 🚫 No need for ETL pipelines
- 🔒 Strong consistency for both workloads
- 💸 Lower infra & ops cost
⚠️ Challenges
- Still a maturing field — not all HTAP systems are equal
- Balancing latency (OLTP) with throughput (OLAP)
- Can be harder to tune for both workloads simultaneously
References
- ChatGPT