Typical Data Profile
Typical users are engineering, production, quality control, and laboratory personnel at printed circuit board factories. Daily documentation includes engineering production instructions, Gerber analysis records, drilling and lamination parameters, chemical solution analysis reports, AOI reports, electrical test reports, and slice inspection reports. Data sources include customer-provided Gerber files, drill tapes, BOMs, CAM files, equipment operating parameters, bath test results, AOI and electrical test data, and laboratory test records. Update frequency follows per production batch, per shift, or continuous collection. Data includes detailed fields such as drilling rotation speed, lamination temperature, and defect coordinates. Units include revolutions per minute, degrees Celsius, millimeters, and others.
Constraints on Deployment and Upgrade
The multi-source heterogeneous data structure and high-frequency update rhythm of PCB manufacturing require deployment workflows to adapt to parsing rules for different document formats. The workflows must support unified access to equipment data, customer files, and laboratory reports. Continuously collected real-time data requires low-latency synchronization links to avoid delays in process parameter queries. Differences in detailed fields across various documents require vector database indexes to support custom field mapping. This ensures key information such as drilling parameters and defect records can be accurately retrieved. The upgrade phase must be compatible with existing historical Gerber and AOI data to prevent format corruption during bulk data migration.
Configuration Settings
This document is compatible with FastGPT v4.17.0.
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
DB_MAX_LINK | 10–15 connections | PCB manufacturing data is multi-source and high-frequency, preventing database connection exhaustion that causes query interruptions |
REDIS_URL | local 6379 port instance | Cache frequently queried process parameters and standard documents to reduce database access pressure |
STORAGE_S3_ENDPOINT | enterprise self-built object storage endpoint | Store large-volume Gerber files, AOI reports, and other production documents to ensure read/write stability |
FE_DOMAIN | factory intranet domain name | Restrict access to internal engineers only, avoiding security risks from public network exposure |
FILE_TOKEN_KEY | randomly generated 32-bit secret key | Encrypt file download links to prevent unauthorized access to production documents |
PG_URL | enterprise self-built PostgreSQL instance address | Store structured metadata for process specifications, enabling fast retrieval by production batch |
Three Common Configuration Mistakes
- Phenomenon: A dependency missing error occurs when running the source code deployment script. Cause: The required version of Node.js and MongoDB dependencies were not installed in advance, and correct environment variables were not configured.
- Phenomenon: The warning
WARNING: current commit information was not captured by theis output during local image packaging. Cause: The packaging command was not executed in the Git repository root directory, or the first Git commit was not completed. - Phenomenon: Q&A results return process parameters that do not match actual production requirements. Cause: Latest data was not synchronized at the production batch or shift update frequency, resulting in expired lamination and etching compensation parameters being stored in the vector database.
How to Verify Successful Configuration
- Run the built-in database connection test tool to verify that the number of connections configured for
DB_MAX_LINKcan be reused normally, with no timeout or connection rejection errors. - Upload a test Gerber file and AOI report to verify that the file upload and parsing process works normally, with no format parsing failure prompts.
- Submit a test query such as "Standard feed rate parameters for drilling processes" to verify that the returned results include correct fields and corresponding values.
- Simulate a production batch data update to verify that the vector database index synchronization is completed automatically, and query results include the latest process parameters.
The values provided are common starting points and should be evaluated against the user's own samples.