DAIV / PRODUCT ARCHITECTURE
Real-DPSDatabase replication & recovery
Real-DPS captures database logs, replicates data and replays changes on standby databases. It manages replication policies, status, verification and controlled failover. Execution engines exchange data through core switches; management interfaces use a separate management switch at each site.
Interactive 3D architecture01 / Product Architecture
Replication built in.
Recovery under control.
Production Database Site, Database DR Platform, Standby Database Site. Select a zone or device to inspect its structure and responsibilities.
- 01Log Capture & Replay
- 02Real-time Replication
- 03Replication Policies & Monitoring
- 04Controlled Failover & Verification
02 / Operating Workflow
Follow the data.
Understand each control.
Explore the Real-DPS workflow step by step. Paths follow the selected step; critical actions retain human confirmation.
This step requires human confirmation. Check rules and permissions before continuing the simulation.
Product simulation only. No production connection or real business operations.
03 / AI-assisted Management
Evidence for decisions.
Boundaries for execution.
Identify issues, explain evidence and suggest actions throughout the database replication & recovery workflow.
Check standby replay and takeover readiness
Explain replication anomalies and propose verifiable actions
AI explanations and advice do not imply automatic execution.
Deterministic rules, permissions and human approval govern high-risk actions.
ARCHITECTURE NOTES
Database replication & recovery,
Clear responsibilities at every layer.
Each site has one core switch and one separate management switch, so the diagram shows one inter-site replication backbone. Dual-core deployments use two redundant replication links. The management network carries policies, status and authorized scheduling, not replicated data. AI assistance does not replace human authorization.
01Production Database Site
Production Database Clusters
Generates business data and logs. The product's production-side engine performs log capture and replication.
Capture & Replication Engine
The Real-DPS production component captures logs, applies replication policies and sends changes through both sites' core switches to the standby engine.
Production Core Switch
Carries the engines' inter-site data traffic. The switch does not perform database replication logic.
Production Management Switch
Connects the platform to device management ports through a separate network; it does not forward business or replication data.
02Database DR Platform
Real-DPS
The product includes replication execution and DR management. Its console configures engines, monitors status and coordinates controlled failover through both management switches. The console does not forward replication traffic.
03Standby Database Site
Standby Database Clusters
Receives changes replayed by the standby engine and maintains a verifiable standby database state.
Receive & Replay Engine
The Real-DPS standby component receives and replays replicated changes into standby databases, reporting progress and execution results.
Replication Verification
Checks lag, log continuity, data consistency and failover results.
Standby Core Switch
Routes incoming inter-site replication data to the receive and replay engine.
Standby Management Switch
Connects the platform to device management ports through a separate network; it does not forward business or replication data.
01 / Configure Replication & Management
Configure capture, receive and replay engines through the management switch at each site. Data cores and management networks remain separate.
02 / Continuous Replication & Replay
The production engine captures changes and sends them through the production and standby cores for standby replay. The platform continuously manages replication tasks and status.
03 / Authorize Controlled Failover
After checking replication status, takeover readiness and permissions, schedule standby database takeover through the management network.
Human authorization required04 / Validate
Verify replication and failover results, report them to the platform and retain evidence. Verification feedback is not an additional replication link.