What data for this category includes
The data for this category comes from the cell grading system, glass adhesive film backplane batch ledger, factory MES system, string welding and lamination production equipment, EL defect detector, and IV power tester. Data updates generate a record for each module, and synchronize batch files for each raw material batch. Daily documents include layout diagrams, material batch records, lamination process parameter records, EL defect images, IV test reports, insulation withstand voltage test records, and finished product packing lists. Core fields include cell current gear, lamination temperature and pressure parameters, EL crack coordinates, IV power values, and more.
What constraints apply during deployment and upgrade
Deployments must configure high-concurrency storage links to handle real-time data updates for each module and prevent data backlog. Upgrades must retain original data source adaptation rules to support multi-source heterogeneous data access and avoid interrupting data synchronization for production links. Configuration items must support associating historical records by module granularity to match the strong binding between core fields and individual modules. Upgrades must be compatible with legacy data field mapping logic. Deployments must configure cross-data-source batch matching rules to support the association between raw material batches and modules, preventing data misalignment during supplier auditing.
Configuration Settings
The following configurations apply to FastGPT v4.17.0. Adjust settings based on the actual factory production environment.
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
DB_MAX_LINK | 100–150 connections | Adapts to high-frequency data write requests generated per module, avoiding connection pool exhaustion |
MONGODB_URI | Fill in according to the production environment cluster address | Stores unstructured data such as EL images and IV test reports, requiring high-availability cluster support |
REDIS_URL | Cluster address with password authentication | Caches component batch matching results and real-time process parameters to reduce repeated calculation overhead |
STORAGE_S3_ENDPOINT | Factory internal object storage node address | Stores batch-generated EL images and packing lists to meet local data compliance requirements |
FILE_TOKEN_KEY | 32-character random string | Encrypts access tokens for stored EL images and packing lists to prevent unauthorized downloads |
ROOT_KEY | Generate random key according to deployment environment | Controls system backend access permissions to avoid unauthorized configuration modifications |
Three common configuration errors
- A Request Timeout error occurs when calling the large language model. This is caused by failure to configure reasonable timeout parameters, leading to link interruption during time-consuming tasks such as EL image interpretation and power matching.
- The function call function cannot be triggered, and a permission exception prompt is displayed. This is caused by incorrect configuration of
ROOT_KEYpermissions, preventing the proxy tool from calling the preset function logic for supplier batch matching. - Empty fields appear during data synchronization. This is caused by failure to adapt to format differences across multiple data sources, leading to incorrect mapping of temperature parameters in lamination process records to system preset fields.
How to verify a successful configuration
- Check the system data source access panel to confirm that all production equipment and MES system connection statuses are normal.
- Submit test data for a single module to verify that EL images and IV test reports can be uploaded normally and associated with the corresponding component file.
- Trigger a supplier batch matching task to confirm that returned results match the business rules for raw material batch ledger matching logic.
- Check the backend running logs to confirm that there are no frequent connection timeouts or data write failure records.
The values given are common starting points and should be measured against the reader's own samples.