Reference only; verify against the target FastGPT release and deployment documentation.
What data looks like for this category
Data related to port shipping document and material drafting primarily comes from four categories: vessel berth declaration forms, container loading and unloading operation slips, port dispatch logs, and goods customs clearance documents. Data update frequency follows the operation workflow. Individual operation documents are generated in real time, while batch dispatch documents are aggregated daily. Individual document lengths range from hundreds to thousands of characters. Fixed fields include vessel name, voyage number, cargo category, cargo tonnage, operation berth, and operation duration. Units include tons, meters, hours, TEU, and others.
Constraints on deployment and upgrade processes
The multi-source, wide length range, and fixed-field characteristics of port shipping document data impose multiple constraints on deployment and upgrade work. Dispersed intranet data sources require configuration of intranet-accessible storage and database connection parameters to avoid cross-network access delays. The wide range of document lengths requires pre-set chunking rules adapted to different document sizes to prevent over-chunking or under-chunking. The large number of fixed fields and diverse units require configuration of field mapping templates during knowledge base initialization to reduce subsequent manual adjustment costs. High-frequency real-time document generation requires adjustment of the database connection pool limit during deployment to ensure stability during concurrent imports.
Configuration settings
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
MONGODB_URI | mongodb://intranet MongoDB address:27017/fastgpt?authSource=admin | Port business deployments run on intranets. This points to an intranet-deployed MongoDB instance to avoid public network access risks |
PG_URL | postgresql://intranet PG address:5432/fastgpt?user=custom username&password=custom password | Stores metadata and session records for port documents. Requires an intranet-accessible PostgreSQL instance |
FE_DOMAIN | https://port intranet domain:3000 | Adapts to the intranet access address after nginx reverse proxy configuration, ensuring internal users can access frontend pages normally |
STORAGE_S3_ENDPOINT | http://intranet MinIO address:9000 or local | Most port document files are stored on intranet storage services. Configure the corresponding intranet storage endpoint to ensure intranet access for file upload and download |
DB_MAX_LINK | 20–30 | Handles high-frequency small-batch document import requests, avoiding database connection failures caused by exhausted connection pools |
FILE_TOKEN_KEY | 32-bit random string generated based on business scenarios | Used to sign file upload tokens, preventing unauthorized intranet file upload operations |
Three common configuration errors
- Issue: A MongoDB connection error appears after starting the container, with logs showing connection timeout. Cause: The intranet address configured in
MONGODB_URIcannot be accessed by the FastGPT container, or the database has not enabled the corresponding port permission. - Issue: When accessing via the intranet domain name configured with nginx reverse proxy, the chat box cannot upload files. Cause:
FILE_TOKEN_KEYis not configured correctly, orFE_DOMAINdoes not match the domain name used for the nginx proxy, leading to failed upload token verification. - Issue: After configuring the Ollama model channel, models cannot be loaded and a connection timeout prompt appears. Cause: The FastGPT container is not deployed on the same intranet network segment as the Ollama container, or a public network address is used to access the Ollama service instead of pointing to a locally or intranet-deployed Ollama instance.
How to verify configuration completion
- Access the system settings page in the FastGPT backend, check the database connection status, and confirm that the instances pointed to by
MONGODB_URIandPG_URLcan connect normally. - Upload a port operation document file, check whether the file can be stored normally and displayed in the knowledge base, and confirm that the
STORAGE_S3_ENDPOINTconfiguration takes effect. - Initiate a batch import request for multiple documents, observe the usage status of the database connection pool, and confirm that the
DB_MAX_LINKconfiguration can support concurrent requests. - Access the configured
FE_DOMAINfrom another intranet terminal, confirm that the frontend page loads normally, and that the chat function operates without issues.
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.