What the data for this category looks like
Daily documents for this category include mold design task sheets, DFM reports, part drawings, machining process cards, trial mold reports, mold modification notices, and mold records. Data sources include customer product models, mold flow or forming simulation files, CAD/CAM output files, machining and measurement data, and trial mold sample data. Data is updated according to project timelines. New records are generated each time a mold is trialed or modified. Data fields include mold number, defect type, mold modification count, machining accuracy parameters, and other fields. Some parameters use industry-standard units such as millimeters and hours.
What constraints these characteristics impose on deployment and upgrade
Mold design and manufacturing data includes large multi-format files and structured records. Deployment must support importing and storing data from multiple sources. The project-based update requirement means upgrade processes cannot interrupt ongoing project data synchronization. Incremental upgrade support is necessary. Mold data from different projects must be isolated independently. Reasonable storage partitions must be configured during deployment. Additionally, mold record data must be retained long-term. The deployed database must support persistent storage to avoid data loss. There are also differing access needs between internal office networks and external networks. Deployment must allow flexible configuration of front-end access domain names.
How to set the configurations
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
MONGODB_URI | mongodb://fastgpt:SecurePass123@db-host:27017/mold_trace?authSource=admin, reserve at least 1000 MB of storage space | Structured data such as mold design records, trial mold reports, and mold modification records require persistent storage to meet long-term data retention needs for multiple projects |
STORAGE_S3_ENDPOINT | http://internal-s3.example.com | Store large-volume files such as CAD drawings, mold flow simulation files, and trial mold sample measurement data. Compatible with S3 protocol internal storage services |
REDIS_URL | redis://:RedisPass456@redis-host:6379/1, set timeout threshold to 600 seconds | Cache conversation context and retrieval results to support real-time queries for mold defect analysis, and avoid repeated calculation of vectors for large-volume documents |
FILE_TOKEN_KEY | 32-bit randomly generated hexadecimal string | Protect access permissions for uploaded mold design files and trial mold data, prevent unauthorized file downloads or tampering |
ROOT_KEY | Strong custom key defined by operations staff | Manage global configurations in the FastGPT backend, limited to core operations staff to reduce the risk of configuration leakage |
FE_DOMAIN | https://fastgpt.mold-enterprise.com | Configure the front-end access domain name to meet access needs for internal office networks or external services of mold enterprises |
Three common mistakes
- Phenomenon: After deploying on Windows systems, model configurations and application lists are empty after device restart. Cause: The FastGPT configuration directory was not mounted to a persistent storage partition. Temporary stored configuration data is automatically cleared after system restart.
- Phenomenon: Service fails to start after modifying
docker-compose.ymlto add environment variables during upgrade. Cause: The environment variable format for FastGPT v4.17.0 was not matched correctly. TheDB_MAX_LINKconfiguration was omitted, or parameter values containing special characters were not wrapped in quotation marks. - Phenomenon: Token consumption details cannot be viewed during conversations, or statistical data is missing. Cause:
REDIS_URLwas not configured to cache session token data, or the log collection switch was not enabled. This prevents real-time calculation of token usage for each conversation.
How to confirm configurations are correct
- Log in to the FastGPT admin backend, use
ROOT_KEYto verify administrator permissions, and confirm that global configuration items can be modified. - Upload a CAD-format mold part drawing, verify that the storage service corresponding to
STORAGE_S3_ENDPOINTcan receive files normally and generate valid access links. - Initiate a simulated defect analysis conversation, check that the database corresponding to
MONGODB_URIcan normally write session records and analysis results. - View monitoring metrics for
REDIS_URL, confirm that the cache service has no timeout errors, and session status can be synchronized normally.
The values provided are common starting points and should be measured against the reader's own samples.