This document is written for FastGPT v4.17.0.
What the data for this category looks like
Daily documents for this category include foundry process cards, batching lists, heat records, mold history records, first article inspection records, material test reports, and nonconforming product review forms. Data sources include ERP material master data, melting furnaces and thermometers, spectrometers, MES heat records, mold management systems, customer drawings and technical specifications. Data updates occur per heat, per batch, or daily. Fields include alloy element content values, melting temperature, pouring parameters, mold usage count, and more. Units include degrees Celsius, seconds, numerical units, and others.
What constraints these characteristics impose on deployment and upgrade
Multi-source heterogeneous data sources require adaptation to different formatted device data and system documents during deployment. Configure adaptation rules for multi-data source connections. High-frequency updated production data requires ensuring synchronization mechanism stability during upgrade. Avoid interrupting real-time synchronization of heat records during version updates. Complex document structures and fields require knowledge base chunking to adapt to long documents and professional terminology. Optimize vector database configuration for foundry defect retrieval needs. Strict quality control requirements require configuring permission tiers during deployment. Ensure access security for process parameters and quality documents.
Configuration Settings
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
DB_MAX_LINK | 10–15 connections | Adapts to concurrent requests for per-heat data synchronization and daily retrieval in foundry workshops. Prevents connection pool exhaustion |
REDIS_URL | cache expiration time 300–600 seconds | Matches the update frequency of heat records. Ensures timeliness of retrieved data while reducing cache refresh overhead |
STORAGE_S3_ENDPOINT | Calibrate based on actual measurements | Requires connecting to the storage path of spectrometer and melting furnace device data. Adapts to device data formats from different manufacturers |
PG_URL | Read-write separated instance | Adapts to high-frequency read-write requirements of ERP and MES. Ensures response speed for supplier audit processes |
ROOT_KEY | Random string of no less than 32 characters | Ensures access permission security for device parameters and quality documents. Meets control requirements of foundry workshops |
Three Common Misconfigurations
- Phenomenon: Knowledge base search returns empty results. Cause:
STORAGE_S3_ENDPOINTis not configured to point to the storage path for spectrometers and heat records. This prevents retrieval of foundry quality-related document data. - Phenomenon: Database connection fails after Docker deployment. Error message:
connection refused. Cause: Connection parameters forPG_URLorMONGODB_URIare not correctly configured. Database service dependencies and port mappings are not declared in docker-compose.yml. - Phenomenon: Admin panel cannot be accessed normally after upgrade. Cause:
FE_DOMAINconfiguration in docker-compose.yml is not updated. Or the old version’sFILE_TOKEN_KEYis not retained, leading to invalid file access permissions.
How to Confirm Configuration is Correct
- Run a connection pool test for
DB_MAX_LINK. Observe connection count changes under concurrent requests. Confirm the value matches actual business concurrent volume. - Upload a foundry heat record document. Run keyword searches. Confirm the corresponding content can be retrieved. Verify correctness of
STORAGE_S3_ENDPOINTand knowledge base configuration. - Restart the Docker service. Access the admin panel interface. Confirm
FE_DOMAINconfiguration takes effect. No permission errors appear. - Check Redis cache logs. Confirm cache expiration time set via
REDIS_URLtriggers refreshes as expected.
The values provided are common starting points. They should be evaluated against the reader’s own samples.