What the data for this category looks like
Business process guidance data for shipbuilding comes primarily from internal manufacturing execution systems (MES), enterprise resource planning systems (ERP), and classification society compliance documents. Update data alongside production batches, and synchronize data per process node during a single project cycle. Document structures include fields such as order number, ship type parameters, process node list, handling materials details, approval flow nodes, and deadlines. Use standard engineering units for fields including deadweight ton, meter, and man-hour. No custom non-standard fields are included.
What constraints do these characteristics impose on deployment and upgrade
Configure cross-system data synchronization links during deployment. Shipbuilding business data comes from multiple systems and updates in batches, so this prevents process guidance failure if a single data source goes offline. Reserve sufficient storage capacity during upgrades. Long production batches generate massive historical documents, so this ensures enough space for stored data. Adjust database connection parameters to adapt to high-volume queries. Preset field mapping rules during deployment. Fields use standardized engineering units, so this ensures parameters displayed in process guidance match on-site construction team habits. Use a rolling update strategy to gradually switch instances during upgrades. This avoids interrupting ongoing production workflows and ensures uninterrupted business operations.
How to set the configurations
Reference only; verify against the target FastGPT release and deployment documentation.
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
MONGODB_URI | mongodb://admin:xxxx@mongodb-host:27017/ship_manufacture?authSource=admin | Shipbuilding business order and ship type archives are stored in MongoDB. Specify the corresponding database and authentication source to ensure normal multi-table queries. |
PG_URL | postgresql://erp_user:xxxx@pg-host:5432/enterprise?schema=production | Internal ERP process node and material data are stored in PostgreSQL. Bind the production schema to isolate non-business data. |
REDIS_URL | redis://:redis_pass@redis-host:6379/1 | Cache user process query context and common process lists. Use an independent database to avoid conflicts with other business caches. |
STORAGE_S3_ENDPOINT | https://s3-cn-south-1.tencentcloudapi.com | Store large-volume business documents such as ship inspection reports and process drawings. Select an S3-compatible storage node matching the deployment region. |
DB_MAX_LINK | 15–25 | Multiple system data sources initiate queries simultaneously. Limit maximum connections to avoid exhausting database resources. The value is determined via actual testing. |
FE_DOMAIN | https://ship-process-guide.example.com | Bind a dedicated access domain name. Ensure internal production personnel can stably access the process guidance page via the enterprise intranet. |
Three Common Misconfigurations
- Symptom:
connection refusederror occurs during docker-compose deployment, with logs showing failure to connect to the MongoDB instance. Cause: The host address and authentication information inMONGODB_URIare not configured correctly, preventing the container from accessing the external database. - Symptom: Process guidance cannot associate context, and the query logic from the previous process node cannot be continued when a user submits a second question. Cause:
REDIS_URLis not configured, or the default shared database is used, resulting in context cache being overwritten or unreadable. - Symptom: Locally modified front-end code cannot be synchronized to the deployed instance, with no changes visible on the page. Cause: Compiled files are not re-uploaded via the static resource path bound by
FE_DOMAIN, or the front-end service is not restarted to load updated content.
How to Confirm the Configuration Is Correct
- Access the domain bound to
FE_DOMAIN, check if the page loads normally with no resource loading failure prompts. - Submit a test business query, verify that the returned results include correct process nodes and field units, with no missing parameters or format errors.
- Initiate two related queries consecutively, confirm that the second query can continue the previous ship type and order information with no context loss.
- Check database connection logs, confirm that the
DB_MAX_LINKconfiguration does not trigger connection limit exceeded alerts, and database resource usage is normal.
The values provided are common starting points and should be measured against the reader's own samples.
Question material comes from public community discussions. Configuration values are common starting points and should be measured against your own samples. Verified on 2026-09-28.