What Data Looks Like for This Category
This category’s data primarily comes from PLM/ERP systems, part inspection data, assembly tool collection terminals, test benches, MES systems, and customer technical agreements. Daily documents include assembly process cards, multi-level BOMs, torque records (unit: N·m), gap adjustment records (unit: mm), coaxiality inspection records (unit: μm), commissioning reports, factory inspection certificates, and packing lists. Data is updated per equipment or order, and includes fields such as order number, equipment model, assembly team, key process parameters, and inspection results.
What Constraints These Characteristics Impose on the Database and Operations Maintenance Link
The need to associate multi-level BOMs and multi-source data requires databases to support nested queries and multi-table join operations to avoid data redundancy. The update frequency per order or equipment requires databases to quickly locate and update single records using unique identifiers such as order numbers and equipment serial numbers. Coexistence of unstructured documents like process cards and reports, and structured torque and gap parameters, requires databases to support both structured storage and unstructured file storage. High-frequency collection of assembly process parameters requires databases to have high-concurrency write capabilities to avoid data loss or delay. The need for data docking with multiple source systems requires databases to support external connections via standard protocols.
How to Set the Configuration
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
MONGODB_URI | mongodb://admin:securepass@db.example.com:27017/assembly_db?authSource=admin&retryWrites=true | Supports mixed storage of unstructured assembly documents and structured process parameters, and adapts to the multi-type data processing logic of FastGPT v4.17.0 |
PG_URL | postgresql://dbuser:dbpass@pg.example.com:5432/assembly_bom | Stores multi-level BOM and order association data, and supports nested queries and multi-table join operations |
DB_MAX_LINK | 18–22 | Matches the high-concurrency write scenario of parameter collection per equipment, and prevents database connection pool exhaustion |
REDIS_URL | redis://:redispass@redis.example.com:6379/0 | Caches frequently queried order and device mapping information to reduce load on the primary database |
| Database connection retry count | 3 times | Adapts to network fluctuations during multi-source system docking, and reduces the probability of data write failures |
Three Common Errors
- Workflow calls to database tools throw "authentication failed" or "connection timeout" errors, or return empty query results. Cause: The username, password, and authentication source parameters in
MONGODB_URIare not filled correctly, or the deployment environment does not open the corresponding port for the database. - Connection wait timeout occurs during high-frequency assembly parameter writes. Cause: The value of
DB_MAX_LINKis lower than the actual concurrent write demand, leading to database connection pool exhaustion. - Multi-level BOM association queries return missing data or disorganized hierarchies. Cause: A database that only supports flat structures is used to store BOM data, and a relational database that supports nested queries via the
PG_URLconfiguration is not used.
How to Confirm the Configuration Is Correct
- Access the database configuration page in the FastGPT backend, test the connectivity of
MONGODB_URI, and verify if a connection success prompt is returned. - Create a test workflow, call the database tool to query a single assembly record, and verify whether the returned fields match the preset equipment parameters and order information.
- Simulate high-frequency write operations, check the occupancy of the database connection pool, and confirm that the value of
DB_MAX_LINKmeets current concurrent demands. - Import a multi-level BOM data set, verify the integrity of the hierarchical structure via association queries, and confirm that the database configuration corresponding to
PG_URLtakes effect.
All listed configuration values are common starting points, and should be evaluated against the reader's own samples.