What data for this category looks like
Mold design and manufacturing production ledger data typically originates from customer product models, mold flow or forming simulation files, CAD/CAM exported drawings, machining and measurement records, and sample test part inspection data. Data updates are triggered by project milestones. New or modified records are added following each mold trial, mold modification, or process adjustment. Daily documents include mold design task sheets, DFM reports, part machining drawings, process cards, mold trial reports, mold modification notices, and mold history records. Fields include mold number, project number, machining parameters, mold trial count, and defect rectification records. Common units are millimeters, hours, piece counts, and similar units.
What constraints do these characteristics impose on deployment and upgrade
Ledger data for mold design and manufacturing includes large CAD and simulation files, and update frequency fluctuates with project iterations. This creates multiple constraints for deployment. First, large file storage requires configuring a storage path adapted for high-capacity files to avoid upload failures. Second, frequent data updates require configuring sufficient concurrent connections for the database connection pool to prevent query timeouts. Third, mixed entry of multiple document types requires configuring general document parsing adaptation rules during deployment to ensure ledger content in different formats is extracted correctly. For the upgrade phase, retain database indexes for historical projects to avoid making old ledgers unsearchable after version updates.
How to set configurations
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
DB_MAX_LINK | 15-20 connections | Mold project data updates frequently; this value balances database load and query response speed |
MONGODB_URI | mongodb://username:password@host:27017/database?authSource=admin | Used to store unstructured documents such as mold trial reports and mold modification notices; must include correct authentication information |
REDIS_URL | redis://:password@host:6379/0 | Caches frequently queried ledger summary results to reduce database pressure |
FILE_TOKEN_KEY | 32-digit random hexadecimal string | Secures download access to uploaded CAD and simulation files |
STORAGE_S3_ENDPOINT | Local storage intranet address or cloud storage service endpoint | Adapts to storage requirements for large mold design files |
This page's configurations apply to FastGPT v4.17.0.
Three common mistakes
- Phenomenon: After deploying FastGPT from source code, connecting to a database deployed via Docker results in an empty knowledge base list, and ledger data cannot be synchronized. Cause: The database address configured in
MONGODB_URIorPG_URLdoes not include correct authentication parameters, or the container network is not connected, preventing access to the database. - Phenomenon: After using Sealos one-click deployment, the interface prompts quota exceeded, and new knowledge bases cannot be created or summary tasks initiated. Cause: The deployment template does not configure sufficient storage space quotas or model call quotas, resulting in insufficient resources.
- Phenomenon: After attempting to configure local ChatTTS to connect to FastGPT, the voice generation function does not respond. Cause:
REDIS_URLis not correctly configured as an intermediate cache channel, causing voice generation requests to fail to transfer normally.
How to confirm configurations are valid
- Run the built-in FastGPT database connection test tool to check if the connection count configured by
DB_MAX_LINKstays within the set range. - Upload a 100MB CAD part drawing to confirm the file upload succeeds and displays normally in the knowledge base retrieval list.
- Initiate a ledger summary task for a mold project to confirm the system automatically pulls associated documents such as mold trial reports and mold modification notices and generates summary content.
- Download the uploaded mold design file to confirm normal access without additional authentication, verifying that the
FILE_TOKEN_KEYconfiguration takes effect.
The values provided are common starting points and should be evaluated against the reader's own samples.