What the data for this category looks like
Daily documents for this category include winding and core assembly records, vacuum drying records, insulating oil reports, partial discharge and withstand voltage test reports, and nameplate parameter tables. Data sources include design BOMs, silicon steel sheet and insulating material data, data collected from vacuum and drying equipment, oil test results, and test station test records. Data is updated once per piece of equipment or batch of materials. Parameters for a single piece of equipment must be fully traceable. Test data from each stage must be tied to the corresponding batch of raw materials. Data formats vary widely across stages, covering structured parameters and unstructured test reports.
What constraints these characteristics impose during deployment and upgrade
Because data must be tied to and updated per piece of equipment or batch of materials, deployment must support configuration for importing multi-source heterogeneous data in batches. The category includes dedicated structured fields such as insulating oil moisture content and withstand voltage values, as well as unstructured test reports. This requires adaptation configuration for both relational databases and object storage. Products require full life cycle traceability, so post-deployment upgrades must maintain compatibility with historical data. This prevents imported test data from becoming unretrievable due to configuration changes. Additionally, some test data involves internal management and control requirements. Deployment must configure parameters related to permission control to ensure the security of data access.
How to set the configurations
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
DB_MAX_LINK | 10–20 | Meets the high-frequency read/write requirements of per-batch updates for single equipment, and prevents database connection exhaustion |
STORAGE_S3_ENDPOINT | Domain name corresponding to the enterprise intranet storage node | Enables fast storage and retrieval of large files such as insulating oil reports and withstand voltage test reports |
REDIS_URL | Connection address of the local Redis instance | Caches historical test data for single equipment, speeding up the retrieval of parameters in tender documents |
MONGODB_URI | Database connection string for mounting design BOMs and test data | Stores structured dedicated category data such as nameplate parameters and partial discharge records |
FILE_TOKEN_KEY | Calibrated through actual testing | Ensures access permissions for confidential test reports in tender documents, preventing unauthorized access |
FE_DOMAIN | Domain name deployed on the enterprise intranet | Adapts to access and configuration adjustments by internal engineers, isolating risks from public network access |
Three common errors
- Pulling the fastgpt-minio image stalls with a connection timeout error. This occurs because no intranet image acceleration address is configured, and
STORAGE_S3_ENDPOINTis not pointed to the local storage node, causing the image pull to rely on public network resources. - After importing CSV data for an insulating oil report, an index error appears after upgrading the version, prompting a field mismatch. This happens because the
DB_MAX_LINKconfiguration value is not adjusted after the upgrade, leading to database connection exhaustion during batch imports and preventing correct parsing of dedicated fields such as insulating oil moisture content and withstand voltage values. - The privately deployed knowledge base interface has no option to add image datasets. This occurs because the permission scope of
FILE_TOKEN_KEYis not correctly configured during deployment, and image format support for the storage service is not enabled, preventing the system from recognizing image files.
How to verify correct configuration
- Run the image pull command, confirm the image is pulled within a reasonable time frame, and check that
STORAGE_S3_ENDPOINTpoints to the intranet storage node. - Import a test CSV file for an insulating oil report, confirm that the database connections
MONGODB_URIandPG_URLcan write data normally, with no field errors. - Enter the knowledge base management interface, confirm that the option to add image datasets exists, and check that the permission configuration for
FILE_TOKEN_KEYcovers image formats. - Run the deployment script on an Arm64 architecture server, confirm that FastGPT v4.17.0 starts normally, and check that the image version is adapted to the target architecture.
The values provided here are common starting points, and should be evaluated against the reader's own samples.