What data looks like for this category
Data for this category comes from ERP material master data, melting furnace thermometers, spectrometers, MES heat records, mold management systems, and customer drawings. Daily documents include foundry process cards, batching lists, heat records, mold histories, first article inspection records, material test reports, and nonconforming product review forms. Fields cover alloy batching ratios, melting temperatures, pouring parameters, defect types, and more. Units include mass percentages, degrees Celsius, seconds, and others. Data update frequency is per heat, per batch, or daily.
What constraints these characteristics impose on deployment and upgrade
This category has scattered data sources with frequent updates. During deployment, connection configurations for multiple heterogeneous data sources must be adapted to avoid data synchronization interruptions. Frequent heat and batch updates require configuring a proper incremental synchronization strategy during deployment to reduce resource consumption from full pulls. Documents contain specialized fields and units. During deployment, the original data structure must be preserved to avoid loss of critical information during parsing. Large volumes of data such as batch-imported heat records require adjusting storage resource configurations during upgrades to prevent import failures or response delays.
How to set configurations
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
MONGODB_URI | mongodb://username:your_password@your_host:27017/fastgpt?authSource=admin | Adapts to storage requirements for structured data such as batch heat records and material test reports |
REDIS_URL | redis://:your_password@your_host:6379/0 | Caches frequently queried process parameters and mold information to reduce database load |
DB_MAX_LINK | 20–30 connections | Adapts to concurrent data access across multiple heats to avoid connection pool exhaustion |
STORAGE_S3_ENDPOINT | https://your_internal_s3_endpoint:9000 | Stores large files such as heat scan records and mold histories |
FILE_TOKEN_KEY | 32-bit randomly generated string | Ensures authentication security for document import and access |
ROOT_KEY | Configure as a fixed key for local deployments | Controls access scope for the platform's highest permissions |
Three common mistakes
- Symptom: Garbled characters appear after importing foundry process documents in a production environment, while imports work normally in a local deployment environment. Cause: The
FILE_TOKEN_KEYwas not configured correctly in the production environment, leading to permission check failure during document parsing and triggering abnormal character encoding. - Symptom: When using FastGPT v4.17.0, responses to process questions of a dozen words first return all results at once before switching to streaming output. Cause: The context cache preheating logic was not adjusted during deployment, causing the first request to generate a full result before enabling streaming transmission.
- Symptom: After upgrading the version, the AI chat node in the workflow cannot obtain the output content of the code running node. Cause: The workflow variable mapping configuration was not updated synchronously during the upgrade, interrupting the data transfer link between nodes.
How to confirm configurations are correct
- Access the platform backend, review data source connection status, confirm normal connection status for
MONGODB_URIandREDIS_URL, and adjust connection count threshold based on actual concurrent volume. - A foundry batching list document may be uploaded to verify that parsed fields are complete and units are not lost, and confirm that the storage path matches the
STORAGE_S3_ENDPOINTconfiguration. - A process parameter query may be initiated to verify that the streaming output logic of responses meets expectations, and check whether the
ROOT_KEYpermission configuration complies with internal management requirements. - A workflow including a code node may be run to verify that the AI chat node can normally obtain the code output content, and confirm that the upgraded variable mapping configuration has taken effect.
The values provided are common starting points and should be measured against the reader's own samples.