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 test sample inspection records. Data updates are carried out on a per-project basis, and are synchronized after each trial mold or mold modification. Documents typically include fields such as mold number, part material, demolding angle, forming parameters, and mold modification records. There are no unified fixed units, and values are adjusted based on document type and enterprise standards.
Constraints Imposed by These Characteristics on Deployment and Upgrade
The data characteristics of this category impose multiple constraints on the deployment and upgrade process. The need to parse multiple types of heterogeneous documents requires adapting preprocessing logic for different formats including CAD, simulation reports, and process cards during deployment. Data updates are triggered by project and trial mold/mold modification nodes, with no fixed cycle. A vector database that supports incremental synchronization must be configured to avoid excessive resource usage from full updates. Cross-departmental cross-review requires associating data from multiple sources, so permission and association rules for different data sources must be integrated during deployment. Frequent mold modification iterations require the upgrade process to ensure the stability of index updates, avoiding interruptions to ongoing review tasks.
How to Configure the Settings
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
REDIS_URL | redis://:custom password@intranet server address:6379/0 | This category has frequent data updates. Redis is used to cache temporarily parsed document fragments and intermediate review results. Database 0 can isolate data from other services |
STORAGE_S3_ENDPOINT | Fill in according to the intranet access address of the enterprise's self-built object storage | Mold design documents are mostly large CAD or simulation files. Intranet storage reduces cross-network transmission latency and costs |
DB_MAX_LINK | 20–30 | A single project cross-review will call multiple document sources and databases at the same time. 20-30 connections balance concurrent performance and server resource usage |
MONGODB_URI | mongodb://username:custom password@intranet server address:27017/fastgpt?authSource=admin | Used for persistent storage of review history, mold modification records, and multi-source document association data. Correct authentication source must be configured |
FE_DOMAIN | Enterprise intranet domain name or public network bound business domain name | Used to generate the front-end access entry for review results, and must match the enterprise's internal access permission rules |
ROOT_KEY | Custom complex string | Used for highest permission verification of the system backend, must match the environment variables of the deployment container |
Three Common Misconfigurations
- Issue: When creating a knowledge base, it is not possible to select a locally deployed large language model, and the interface dropdown list is empty. Cause:
ROOT_KEYand model access permission verification are not configured correctly, or the OneAPI proxy address is not added in the system configuration. - Issue: A simple application operation error occurs when deploying with Sealos, and the interface returns a 500 status code. Cause: The connection parameters of
MONGODB_URIorREDIS_URLare configured incorrectly, causing database connection failure. - Issue: CosyVoice2-0.5B cannot generate speech normally, while whisper-large-v3-turbo works normally. Cause: The voice model call rules for FastGPT v4.17.0 are not adapted, or the voice model's container port is not mapped to an intranet-accessible address.
How to Confirm Successful Configuration
- Log in to the FastGPT system backend, enter the database configuration page, run connection tests for
MONGODB_URIandREDIS_URL, and confirm that a successful status is returned. - Upload a mold design task sheet PDF document, test the document parsing function, and confirm that text and structured fields in the document can be extracted normally.
- Enter the model management page, try to add a locally deployed chatglm2 model accessed via OneAPI, and confirm that the model can be selected in the knowledge base configuration dropdown list.
- Initiate a multi-document cross-review task, confirm that the task can start normally and generate expected review results.
The values provided are common starting points and should be evaluated against the reader's own samples.