What the data for this category looks like
Bid and quotation data for photovoltaic module encapsulation comes from daily output of the string soldering, lamination, framing, and testing departments. Core documents include component layout diagrams, material batch records, lamination process parameter logs, EL defect images, IV power test reports, insulation voltage withstand test records, and finished product packing lists. Data is sourced from battery cell grading ledgers, glass adhesive film backboard batch archives, production MES systems, string soldering and lamination device logs, and EL/IV tester output results. Updates are generated per individual module, with batch archives created per batch of raw and auxiliary materials.
What constraints these characteristics impose during deployment and upgrade
This category’s data has a high update frequency, and single records are linked to multiple document types. During deployment, ensure database connection stability to avoid connection exhaustion from high-frequency read and write operations. Storage and reading of multiple types of heterogeneous documents — including images, structured tables, and process text — require configuring a file storage service that supports multiple formats. Bid scenarios require fast retrieval of historical batch and component data, so reserve sufficient cache resources to support high-frequency queries. During upgrades, maintain compatibility with original data source connections for MES and testing devices to avoid interrupting production-related data synchronization.
How to set configurations
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
DB_MAX_LINK | 100–150 | Adapts to high-frequency read/write demands for component and batch data, preventing service freezes caused by connection exhaustion |
REDIS_URL | Local Redis instance, memory allocated 2000 MB | Supports high-frequency caching of historical bid data and component parameters, reducing database read pressure |
STORAGE_S3_ENDPOINT | Object storage endpoint compatible with multiple formats | Stores multiple types of heterogeneous documents including EL images and IV test reports, matching business data storage requirements |
MONGODB_URI | Independent MongoDB instance, connection parameters calibrated via actual testing | Stores unstructured process records and test data, adapting to high-frequency update business scenarios |
FILE_TOKEN_KEY | Custom high-strength encryption key | Secures access to sensitive bid-related documents, preventing unauthorized retrieval |
ROOT_KEY | Secure key only used for initial configuration | Restricts access to platform initialization operations, preventing unauthorized modification of core configurations |
Three common errors
- Phenomenon: Service remains unresponsive for long periods after deployment startup, or AI chat response times exceed 40 seconds. Cause: Memory allocation for
DB_MAX_LINKandREDIS_URLis not properly configured, leading to resource exhaustion from high-frequency read/write operations. - Phenomenon: Existing bid data cannot be loaded normally after migrating from Sealos environment to local deployment. Cause: Database backups associated with
MONGODB_URIwere not properly exported and restored, and batch and component associated data was omitted. - Phenomenon: AI chat cannot be initiated normally after Docker deployment, returning 400 or 500 status codes. Cause:
FILE_TOKEN_KEYorSTORAGE_S3_ENDPOINTwas not configured correctly, causing document reading failures that trigger error reports.
How to confirm configurations are correct
- Initial configuration verification is performed using the FastGPT v4.17.0 deployment tool, confirming that parameter values for core configuration items including
DB_MAX_LINKandREDIS_URLmatch the preset plan. - One EL test image and one IV test report are uploaded to verify that the file storage and reading process works normally, with no error messages returned.
- One simulated bid document generation request is submitted, confirming that response times meet expected thresholds for the business scenario.
- A database backup file is exported, verifying that the backup content includes all component and batch associated data and can be restored normally.
The values provided here are common starting points, and should be evaluated against the reader’s own samples.