What the Data for This Category Looks Like
Data sources include engineering orders, sand, gravel, cement admixture inspection reports, aggregate moisture content measurement records, mixing station control system logs, laboratory test block strength reports, and similar materials. Daily documents include mix proportion notices, opening identification forms, production and transport orders, raw material inspection documents, test block records, and strength reports. Data updates follow per vehicle production record, per batch raw material inspection result, and daily summary report. Document fields include raw material moisture content, mixing duration, slump value, and test block compressive strength value, with corresponding units of percentage, seconds, millimeters, and megapascals.
Constraints Imposed by These Characteristics During Deployment and Upgrade
High-frequency per-vehicle production data and daily batch summary reports require sufficient database connection pool and storage capacity during deployment. This avoids data write blocking or insufficient storage space. Differences in structured fields across various documents require configuring appropriate document parsing rules during deployment. This ensures the equipment inspection and maintenance Q&A knowledge base can accurately extract key information such as raw material parameters and production records. Frequent updates to raw material inspection data require ensuring uninterrupted data synchronization during upgrades. This prevents outdated information from appearing in Q&A results. Restricted network environments may exist inside mixing stations. Network proxy and mirror source configurations must be adapted during deployment, to ensure containers can pull images normally.
How to Set the Configurations
This configuration applies to FastGPT v4.17.0.
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
DB_MAX_LINK | 50–80 connections | Adapts to the needs of high-frequency per-vehicle data writes and concurrent queries across multiple departments |
MONGODB_URI | mongodb://fastgpt:securepass@mongodb:27017/concrete_check?authSource=admin | Stores unstructured inspection logs, test block reports, and production records, supports flexible document structures |
PG_URL | postgresql://fastgpt:securepass@postgres:5432/concrete_check | Stores structured mix proportion, production order, and raw material inspection data, ensures transaction consistency |
REDIS_URL | redis://:securepass@redis:6379/0 | Caches frequently queried raw material parameters and inspection rules, reduces database load |
FE_DOMAIN | https://concrete-check.your-domain.com | Binds a dedicated domain, adapts to access via the internal office network of mixing stations |
Three Common Configuration Mistakes
- Symptom: The knowledge base cannot recall historical raw material inspection data for ready-mixed concrete, and query results are empty. Cause: Incorrectly used
PG_URLinstead of vector database configuration, and failed to configureMONGODB_URIor vector database connection parameters. This results in inability to store and recall unstructured inspection and test data. - Symptom: Docker image pull times out, with an error showing network connection failure. Cause: Failed to configure proxy or mirror source adapted to the restricted internal network of the mixing station. This results in inability to pull FastGPT-related container images.
- Symptom: Q&A results show outdated aggregate moisture content data. Cause: Did not pause the data synchronization script during upgrade, leading to data synchronization conflict between the old and new version knowledge bases. This overwrites the latest inspection records.
How to Confirm Proper Configuration
- Access the system settings page in the FastGPT backend, check the database connection configuration panel, and confirm that the configuration values of
DB_MAX_LINK,MONGODB_URI,PG_URL, andREDIS_URLmatch the preset values. - Upload a real mix proportion notice document, and verify that the knowledge base can correctly extract key fields such as raw material names, moisture content, and mix proportion values.
- Initiate a query related to equipment inspection, and confirm that the recalled results include the latest production transport order and test block strength report data.
- Check container runtime logs for no error messages related to database connection timeouts, image pull failures, or data synchronization exceptions.
The values provided are common starting points, and should be evaluated against local samples.