A time-series database platform with a Rust core, designed for high-frequency time-series ingestion, query, management and operations. It forms a complete engineering stack of database core + routing layer + management plane + multi-language SDKs + operations scripts.
CberaRTDB is not a single embedded library, but a set of time-series data platform components that can be deployed independently, scaled horizontally, accessed by multiple clients and managed through a web console. Its goal is to give business systems unified time-series ingestion and query capabilities, to provide a unified routing entry point for multi-instance, multi-shard and multi-replica scenarios, and to give operators and administrators management, auditing, export, monitoring and configuration visibility.
The current version v1.3 already includes key deliverables such as raft_store persistence, 24h soak testing and 6 of 7 items checked on the overall acceptance checklist.
The core execution engine, handling TCP protocol processing, account authentication, SQL V2 execution, WAL/Segment persistence, flush/compaction, replication, migration and Prometheus metric exposure.
A unified client access entry point that forwards requests to the primary shard per configuration. Supports static sharding and consistent hashing, ring version migration, config fingerprint validation, micro-batching and connection pooling.
A Rust-implemented REST API gateway that wraps the database TCP capabilities, exposing users, databases, tables, SQL, tasks, auditing, export and monitoring to the web console.
Built with React + TypeScript + Vite + Ant Design + Monaco, providing login, dashboard, user/database/table management, a SQL workbench, runtime monitoring, background tasks and audit log pages.
The Python SDK suits scripting and data processing; the .NET SDK suits enterprise applications, desktop tools and server-side systems. Both support an adaptive batch writer, pipelining and connection reuse.
Supports generating binaries via cargo build --release, one-click systemd and Windows Service setup, CI workflows, stress-test and replay scripts, plus canary releases and backup/restore.
Login, query, write and non-query commands over a custom TCP protocol; plain and batch writes; SQL V2 system table queries; WAL/Segment/flush/compaction background mechanisms.
Router sharded routing (static ring / consistent hashing); Raft cluster consensus from 3 nodes with 30s RTO leader election; --no-sql hot-path stripping; BlockExecutor with zero SQL AST dependency; ring expansion migration; full backup and restore.
Prometheus metric endpoint (including Raft role/term/commit/applied metrics); Grafana dashboard (5 rows, 14 panels); 11 P1/P2/P3 graded alert rules; W3C TraceContext + OpenTelemetry; Python SDK ClientMetrics.
Queries SELECT/EXPLAIN; INSERT/UPDATE/DELETE; CREATE/DROP DATABASE/TABLE; SHOW TABLES/DESCRIBE; ADD/ALTER/DROP USER. Expressions support BETWEEN/LIKE/IN/IS NULL and COUNT/MIN/MAX/SUM/AVG aggregates.
Admin Gateway wraps the REST API; Admin Web delivers login, dashboard, user/database/table management, a SQL workbench, runtime monitoring, background tasks and audit logs; SQL history, favorites, export and persisted auditing.
The Python SDK supports connection, authentication, query, write and batch write; the .NET SDK supports connection strings, synchronous calls and P4.1 multiplexing. Bundled integration and stress test tools simplify joint debugging.
For industrial and business teams that need a self-controlled time-series data foundation, usable as a private data platform, an equipment-data foundation or a domestic technology substitution.
Storage and query of time-series data from industrial equipment, sensors and edge acquisition.
Business runtime metrics, operational log summaries and event stream archiving.
Business systems that need high-frequency writes and time-range queries through an SDK.
Building internal data platforms that need a lightweight management plane and auditing capabilities.
A manufacturing execution system purpose-built for modern automated quenching and tempering lines. It covers the full closed loop from plan intake and intelligent scheduling to quality traceability, making production transparent, controllable and traceable.
The project vision is to bridge the information gap between the planning layer (ERP / MES) and the control layer (PLC / DCS), eliminating data silos and achieving fully digital, intelligent control from order release to finished-goods warehousing.
The system adopts a modern layered architecture split into presentation, business, integration and data layers. The business layer builds core services for planning, scheduling, tray loading, recipes, monitoring, quality, equipment and traceability, which together form the core capabilities of the MES.
Web console, shop-floor touch terminals and production dashboard screens provide multi-device coverage for real-time visibility of progress, equipment status and quality metrics.
Hosts planning and scheduling, tray loading execution, process recipes, progress monitoring, quality control, equipment O&M, full-process traceability and reporting analytics.
Standardized adapters connect to PLC, SCADA and shop-floor MES; message queues plus task scheduling break down information silos and ensure data consistency.
Business database, time-series/snapshot storage, log and audit storage and high-speed caching provide a secure, reliable, high-concurrency production data foundation.
Centered on the production work order, connecting the full chain from plan intake and intelligent scheduling to finished-goods output.
The shop-floor MES issues production work order instructions.
The line automatically schedules device-level tasks.
Terminals guide operators to match materials precisely.
Optimal process parameters are matched and issued automatically.
Process data is collected and equipment status monitored.
Quality data is entered and automatically judged by the system.
Exceptions are logged, tracked and closed out.
Production results are aggregated and returned to upper systems.
A modular core service design that adapts flexibly to line upgrades and business iteration.
Work order intake → intelligent scheduling (APS algorithms) → manual adjustment → task dispatch, forming a plan-execute-feedback loop. Supports rush orders, inserted orders and quick rescheduling of changed work orders.
Automatically generates loading plans from BOM and loading rules; terminals, PDAs and electronic kanban guide picking and barcode verification; material batches, pallets and machine positions are linked automatically.
Dual-source collection → intelligent matching → permission validation → automatic distribution → execution feedback. Supports offline editing, online activation and version rollback, with parameter interlock checks to eliminate human deviation.
Millisecond-level collection of equipment status and process parameters feeds multi-dimensional visual dashboards; multi-level alarm linkage reaches owners via SMS, email and audible/visual signals.
The system automatically dispatches inspection tasks and applies national and enterprise standard rules; non-conforming products are graded for disposition; all information is consolidated into a tamper-proof electronic quality record.
Full equipment lifecycle records plus preventive maintenance. Breakpoint resume restores production with one click after a fault is cleared, across shifts and equipment, safeguarding production continuity.
Seamless integration with the dispatch system: production tasks trigger automatic AGV instruction dispatch with real-time status feedback, precisely matching production and logistics takt.
Multi-dimensional precise queries by work order, pallet barcode, workpiece number, equipment task and time range. One click restores the entire production history into a tamper-proof digital record.
Recipes and process parameters are distributed and verified end to end, eliminating manual deviation.
Prevention, in-process control and after-the-fact traceability effectively reduce production defect rates.
Multi-level alarms and closed-loop handling greatly reduce scrap losses caused by parameter drift.
Automatic production data collection and report generation improve management efficiency.
The project has completed all key deliverables for the requirements analysis and detailed design phases as planned, with core development and testing progressing on schedule.
A standardized layered architecture including presentation, business logic, integration and data access layers, with each layer decoupled.
Work order intake and parsing, intelligent scheduling and dispatch, automatic tray loading execution, production progress tracking and other key steps.
Covering work orders, equipment tasks, process recipes, material batches and quality standards, accurately mapping production business entities.
Including the work order header table mes_order_header, the equipment task table mes_device_task, the recipe parameter table and more.
Interface 07 for work order dispatch, interface 08 for tray loading plan dispatch, production completion status feedback and more, connecting upstream and downstream systems.
Query response time <2 s, complex report generation <10 s, MES real-time interaction latency <500 ms.
Whether it is a CberaRTDB foundation partnership or a delivery of the Automated Quenching & Tempering Line MES, we provide full-cycle services from architecture consulting and custom development to operations assurance.
Business Enquiry