What data for this category looks like
This category of data is generated from PLM, ERP, MES, supplier materials, validation test results, and customer-specific requirements. Updates are triggered irregularly with each process or engineering change. Daily documents include ECN/ECR files, change assessment reports, drawings and BOM version records, process validation reports, production switching plans, inventory disposal documents, and customer approval documents. The data links to change workflows across R&D, process, quality, production, and procurement teams. It includes core fields such as change number, version identifier, approval nodes, associated materials and process entries.
What constraints do these characteristics impose on the deployment and upgrade phase
The multi-source and decentralized nature of data for this category requires deployment personnel to configure multi-system docking adaptation rules during the deployment phase. This ensures unified format synchronization for data sources such as PLM and ERP. The irregular update rhythm of changes requires deployment personnel to retain version traceability capabilities for historical change data during upgrades. This avoids data loss. The diverse document types include unstructured drawings and tables. Deployment personnel must reserve storage and parsing configurations compatible with multi-format files. Sensitive data linked to customer approvals and validation tests requires deployment personnel to configure strict permission access rules during the deployment phase. This ensures data access complies with workflow approval requirements.
How to Configure the Settings
| Configuration Item | Recommended Setting | Rationale for This Setting |
|---|---|---|
MONGODB_URI | For local deployment, enter mongodb://localhost:27017/fastgpt; for cloud deployment, enter the corresponding cloud database connection string | This category requires storage of large volumes of unstructured change documents such as ECN/ECR and drawings. MongoDB adapts to multi-format data storage needs |
REDIS_URL | For local deployment, enter redis://localhost:6379/0; set the timeout configuration to 600–1200 seconds | Change workflows require caching approval node status and temporary files. The timeout configuration must match the average time required for cross-team approvals |
STORAGE_S3_ENDPOINT | For local deployment, use the MinIO intranet address such as http://192.168.1.100:9000; for cloud deployment, enter the corresponding cloud storage endpoint | This category requires storage of large volumes of large-volume files such as drawings and validation reports. S3-compatible storage adapts to multi-scenario deployment needs |
FILE_TOKEN_KEY | Use a random secret string of at least 32 characters in length | Change files require token verification for access permissions. This prevents unauthorized downloading of sensitive approval documents |
DB_MAX_LINK | Configure as 20–50 concurrent connections | Multi-system docking for change workflows requires stable database connections. Excessively high connection counts will lead to excessive resource usage |
Three Common Misconfigurations
- When deploying on Alibaba Cloud, a 400 error occurs when calling the model. The cause is that the permission whitelist for
MONGODB_URIand the cloud database was not correctly configured, resulting in database connection timeout or insufficient permissions. - When uploading change drawings via the API, the agent cannot access the files, and the service becomes unresponsive. The cause is that the correct storage path for
STORAGE_S3_ENDPOINTwas not configured, resulting in file storage path conflicts or missing permissions. - After deploying with docker-compose for a period of time, the workflow and knowledge base pages are blank, but the interfaces can be called normally. The cause is that persistent storage was not configured for
REDIS_URL, resulting in loss of cached page configuration data after restart.
How to Confirm the Configuration Is Complete
- Perform a database connection test to verify that the configurations for
MONGODB_URIandREDIS_URLcan establish normal connections. - Upload a change document in ECN format to verify that the file can be stored normally and accessed via intelligent retrieval.
- Configure cross-team permission rules to verify that users with different roles cannot access unauthorized change approval documents.
- After upgrading to FastGPT v4.17.0, check that no historical change data is lost or has format abnormalities.
The values provided here are common starting points, and should be measured against the reader's own samples.