What the data for this category looks like
Shipbuilding ledger data originates from business tools including production execution systems, material requirements planning systems, hull block quality inspection systems, and machinery installation ledger systems. Data updates are triggered by production milestones during ship construction. Ledgers for a single vessel cover full-process records including block construction, outfitting, and sea trials. Each ledger document includes fields such as vessel name, block number, material code, inspection item, measured value, man-hours, and responsible person. Units mostly use tons, meters, hours, and pieces. The core of each document is a structured table, with unstructured attachments such as quality inspection photos and construction drawings.
What constraints do these characteristics impose during deployment and upgrade
The data volume of shipbuilding ledgers grows continuously with each construction batch. A single knowledge base may hold tens of thousands of structured records and a large number of attachments. Deployment must adapt to high-concurrency database connection and large-file storage requirements. Data update frequencies range from real-time to daily. During upgrades, consistency of ledger indexes must be maintained to avoid field mismatch issues. Large-volume attachments and batch documents extend processing time for indexing and deletion, creating clear requirements for timeout configurations during deployment. Sufficient storage capacity must also be reserved to handle future new ledgers.
How to set configurations
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
DB_MAX_LINK | 100–150 connections | Shipbuilding ledgers have high concurrent query and index request volumes. This parameter prevents database connection exhaustion |
MONGODB_URI | mongodb://{user}:{pass}@{replset-host}:27017/ship_ledger?replicaSet=rs0 | Unstructured attachment metadata and historical ledger records are suitable for storage using a MongoDB replica set to ensure high availability |
PG_URL | postgresql://{user}:{pass}@{pg-primary}:5432/ship_report?sslmode=require | Structured ledger field queries and statistical summaries are suitable for use with PostgreSQL. Primary-secondary configuration improves query stability |
REDIS_URL | redis://{redis-cluster}:6379/0 | Caches ledger metadata and access tokens to reduce repeated query pressure on databases |
STORAGE_S3_ENDPOINT | https://s3.enterprise-shipyard.com | Stores large-volume attachments such as quality inspection photos and construction drawings attached to ledgers, and supports enterprise-built object storage |
FILE_TOKEN_KEY | 32-bit random encrypted string | Protects access permissions for ledger attachments and prevents unauthorized file downloads |
Three common mistakes
- Issue: A
timeout of 60000ms exceedederror occurs when deleting a shipbuilding block ledger folder with a large number of documents in the knowledge base. Cause: The default processing timeout setting is insufficient to complete the batch deletion operation, as the single folder contains thousands of block quality inspection and material requisition documents. - Issue: The ledger database cannot be connected after starting the local development deployment environment. Cause:
MONGODB_URIandPG_URLare not configured to use the dedicated database instance for shipbuilding ledgers, and the default test database address is used incorrectly. - Issue: A service error occurs after upgrading the public cloud version, and ledger data cannot be queried. Cause: The
DB_MAX_LINKparameter is not adjusted for the high-concurrency ledger query scenario, and the configuration file is not backed up in advance, making rapid downgrade and rollback impossible.
How to confirm the configuration is complete
Reference only; verify against the target FastGPT release and deployment documentation.
- Upload a test folder containing more than 1000 shipbuilding ledger entries, wait for indexing to complete, and check that the number of documents in the knowledge base matches the actual uploaded number, with no timeout errors.
- Access the management backend at
FE_DOMAIN, generate an access token for a ledger file, and confirm that the test attachment can be downloaded normally using the token. - Adjust the
DB_MAX_LINKparameter, then use a database monitoring tool to check the connection count. Confirm that the connection count stays within the configured range, with no peak overflow.
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.