What This Category of Data Looks Like
Data covers end-to-end records from ironmaking, steelmaking, continuous casting, quality control, and scheduling departments of integrated iron and steel enterprises. Daily documents include heat plans, batching lists, converter smelting records, composition reports, continuous casting process records, and slab judgment sheets. Fields include heat number, hot metal ratio, smelting duration, composition test values, casting speed, and more. Units include degrees Celsius, percentage, and meters per minute. Data sources include hot metal and scrap data, L2 process control systems, laboratories, MES, energy medium systems, and equipment status. Update frequencies fall into three categories: per heat, real-time, and per shift.
Constraints Imposed on Deployment and Upgrade
The high concurrency, multiple data sources, and frequent updates of this category impose multiple constraints on deployment and upgrade. Multi-source data docking requires configuring multiple storage connections to avoid cross-system call conflicts. Real-time and per-heat update frequencies require reserving sufficient capacity for database connection pools. This prevents process interruptions caused by exhausted connections. High-temperature continuous production requires using a gray release strategy during upgrades. This avoids downtime that affects coupled processes for heats and casting sequences. Large-volume process records and device logs require configuring parameters adapted to large-capacity storage. This ensures data persistence.
How to Set Configurations
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
DB_MAX_LINK | 100–150 connections | Adapts to high-frequency data writes per heat and in real time, avoiding connection exhaustion during multi-source data docking |
MONGODB_URI | mongodb://admin:password@internal-mongo:27017/fastgpt?authSource=admin | Stores unstructured documents such as composition reports and continuous casting process records, adapting to large-volume data |
REDIS_URL | redis://:redis-pass@internal-redis:6379/0 | Caches real-time heat status data, reducing query pressure on the primary database |
STORAGE_S3_ENDPOINT | http://internal-oss:9000 | Connects to the enterprise’s internal object storage for storing large files such as smelting process videos and equipment status logs |
FE_DOMAIN | https://steel-audit.internal.com | Configures the access domain name of the enterprise’s internal audit system to enable jump docking with office systems |
Three Common Mistakes
- Phenomenon: The locally deployed workflow configuration interface has text input lag when displaying many nodes. Cause: The
DB_MAX_LINKparameter is not adjusted. Insufficient database connection pool capacity causes cache read delays and interface rendering lag. - Phenomenon: A database connection failure error appears after running the
docker compose pullupgrade command. Cause: Connection permission configurations forPG_URLorMONGODB_URIare not updated. The new version of the image cannot access the original data storage service. - Phenomenon: Custom audit templates cannot be imported during private deployment. Cause: The
FILE_TOKEN_KEYparameter is not configured. This causes permission verification failures for template files, and the interface blocks import requests.
How to Confirm Correct Configuration
- Log in to the data source management page in the FastGPT v4.17.0 backend. Check the service connection status for
MONGODB_URI,PG_URL, andREDIS_URLone by one. Confirm there are no error messages. - Simulate generating a heat record to trigger the supplier and process quality audit workflow. Confirm that data writes to storage normally, and that correct field information including heat number and composition test values is returned.
- Access the configured
FE_DOMAINdomain name. Confirm the interface loads normally, and that clicking the internal system jump button connects properly to the enterprise office system. - Run the
docker compose up -dcommand to restart services. Check container logs. Confirm there are no error messages related to database connection failures or insufficient storage permissions.
The values provided are common starting points and should be measured against the reader's own samples.