What data falls under this category
Data for this category originates from manufacturing execution systems (MES), online inspection equipment, laboratory instruments, process parameter logs, customer quality standards, and historical defect databases. Collection runs continuously per production batch and shift. Daily documents include control plans, patrol inspection records, SPC statistical reports, final product inspection reports, nonconforming product review forms, release approval forms, and monthly quality summary reports. Data fields cover inspection batch identifiers, nonconforming item codes, defect levels, process parameter thresholds, and release determination results. No mandatory fixed units apply; values adjust flexibly based on document type.
What constraints this category imposes on deployment and upgrades
This category’s multi-source heterogeneous data sources, high-frequency update cadence, and varied document structures create multiple constraints for deployment and upgrade workflows. Adapt to access protocols for different data sources including MES and inspection equipment, and reserve configurable data source interface mapping entries. Batch-level and real-time data collected continuously require database and cache service connection parameters to align with throughput demands. Differentiated document structures such as control plans and final product inspection reports require preset, adjustable parsing rule templates. Release review permissions are tied to enterprise organizational structures, requiring synchronized configuration of identity verification associated mapping entries. During upgrades, maintain associated integrity between historical inspection batches and defect data to prevent data disruptions that impact audit workflows.
How to configure settings
The following configurations apply to FastGPT v4.17.0.
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
MONGODB_URI | mongodb://username:password@intranet host IP:27017/fastgpt_qc?authSource=admin | Stores historical defect databases, inspection reports and other quality data; requires intranet authentication and target database designation |
REDIS_URL | redis://:custom password@intranet host IP:6379/0 | Caches frequently queried SPC reports and release approval cache data; requires authentication to ensure data security |
DB_MAX_LINK | 20–30 | Accommodates concurrent requests for multi-source data generated per production batch, prevents service interruptions caused by exhausted connections |
STORAGE_S3_ENDPOINT | http://intranet storage node IP:9000 | Stores attachments including patrol inspection records and final product inspection reports, meets private storage requirements for intranet deployments |
FE_DOMAIN | http://intranet deployment server IP:3000 | Configures frontend access addresses, supports local access scenarios for intranet environments |
ROOT_KEY | Randomly generated 32-character alphanumeric string | Configures the system’s highest privilege access key, ensures initial permission security after deployment |
Three common configuration errors
- Phenomenon: External model services cannot be accessed during intranet deployment. Logs show connection timeout or 407 proxy authentication failure. Cause: HTTP proxy parameters with username and password are not configured, or the proxy configuration format does not include authentication information.
- Phenomenon: Custom models configured for simple applications become invalid after an upgrade. Chat automatically switches to the default model. Cause: Associated mapping items for model configuration were not migrated correctly during upgrade, or configuration files were overwritten, resulting in loss of custom configurations.
- Phenomenon: A database connection failure prompt appears when accessing the workflow. Cause: The
DB_MAX_LINKconfiguration value is too low, or authentication information inMONGODB_URIis entered incorrectly. This leads to exhausted concurrent connections or authentication failures.
How to verify successful configuration
- Upload a final product inspection report, confirm the system correctly parses core fields including inspection batch and nonconforming items, with no parsing failure prompts.
- Initiate a simulated release review workflow, confirm permission configurations are correct, and approval nodes receive task notifications normally.
- View the database connection monitoring panel, confirm current active connections do not exceed the
DB_MAX_LINKconfiguration value, with no connection timeout errors. - Test the intranet proxy configuration, access the external model service interface, and confirm normal return of response data.
The values listed here are common starting points, and should be assessed against the user’s own operational samples.