What the data for this category looks like
Production data for this category comes from formula systems, raw material batch ledgers, internal mixers, Mooney rheometers, vulcanizing machines, and laboratory testing equipment. Daily generated documents include formula sheets, weighing records, mixing curves, rubber compound quick test reports, vulcanization process cards, mold usage records, and finished product performance reports. Data fields include raw material batch numbers, mixing temperatures, vulcanization durations, finished product tensile strength, and more. Units include degrees Celsius, minutes, MPa, and others. Data is updated for each batch of rubber, each batch of raw materials, and each molded product. Single data volume ranges from tens of KB to several MB, and overall data volume grows linearly with production scale.
What constraints these characteristics impose during deployment and upgrade
High-frequency single-data updates require deploying storage connections configured to support high-concurrency read/write operations, to avoid data synchronization delays that affect ledger real-time performance. Multi-source heterogeneous data sources require preset data mapping rules, to ensure log formats from different devices are unified before connecting to the ledger system, reducing subsequent parsing and summary workload. Diverse document types require adapting to different file parsing formats, to avoid parsing failures caused by format differences between formula sheets, mixing curves, and other files. At the same time, errors in this category’s data can lead to product scrap or unqualified performance. Deployment requires configuring permission control mechanisms to prevent unauthorized modification of ledger content. Upgrades require retaining the immutability of historical data, to ensure the accuracy of production traceability.
How to set the configurations
The following configurations apply to FastGPT v4.17.0, optimized for the production ledger summary and report generation scenario for rubber mixing and product vulcanization.
| Configuration Item | Recommended Value | Rationale for This Value |
|---|---|---|
MONGODB_URI | mongodb://admin:securepass@192.168.1.10:27017,192.168.1.11:27017/fastgpt_manufacture?authSource=admin | This category has frequent data updates. The cluster-mode MONGODB_URI supports high-concurrency ledger read/write operations and index construction |
REDIS_URL | redis://:redispass@cache-host:6379/0?maxmemory-policy=allkeys-lru&maxmemory=1000 MB | Caches frequently accessed mixing curves and rubber compound quick test reports. The 1000 MB memory configuration meets the caching needs of single-batch production data |
STORAGE_S3_ENDPOINT | http://s3-cn-north-1.example.com | Stores large-volume documents such as batch-uploaded mixing curve PDFs and finished product performance reports, adapting to multi-file concurrent storage requirements |
FILE_TOKEN_KEY | 32-character random string | Protects sensitive documents including uploaded formula sheets and process cards, preventing unauthorized access and tampering |
FE_DOMAIN | https://manufacture-doc.example.com | Configures the frontend access domain name, used to generate shared links for ledger summary reports and frontend page loading |
ROOT_KEY | 64-character random secret key | Configures system administrator permissions, ensuring only authorized personnel can modify configuration rules for ledger summary and report generation |
DB_MAX_LINK | 20 | Configures the database connection pool size. The value of 20 meets the concurrent read/write needs of this category’s high-frequency data |
Three common mistakes
- A
400 Bad Requesterror is returned after uploading a mixing curve PDF or formula sheet, indicating file parsing failure. The cause is thatFILE_TOKEN_KEYis not configured, leading to failed file upload permission verification, preventing file synchronization to the storage service corresponding toSTORAGE_S3_ENDPOINT. - New users cannot complete registration after deployment, with the interface prompting "Registration interface not enabled". The cause is that the user registration-related configuration switch is not enabled during open-source deployment, and system permission-related parameters are not correctly configured.
- Knowledge base retrieval cannot match weighing records for a specified batch, with the number of retrieval results far fewer than the actual number of ledger entries. The cause is that
DB_MAX_LINKis set too small, exhausting the database connection pool, leading to partial batch data not being synchronized to the knowledge base index.
How to confirm configurations are correct
- Access the frontend address corresponding to
FE_DOMAIN, confirm that the ledger summary page loads normally, with no loading failures or resource 404 errors. - Upload a real mixing curve PDF file, confirm that mixing temperature, duration, and other fields are correctly extracted after parsing, with no empty or garbled data.
- Run a ledger generation task, confirm that the generated report includes corresponding fields from formula sheets, weighing records, and vulcanization process cards, with data consistent with source documents.
- Check the connection logs for
REDIS_URL, confirm no continuous connection timeouts or authentication failure errors, with the caching service running normally. - Check the database connection logs for
MONGODB_URI, confirm no connection failures or authentication errors, with ledger data synchronization working normally.
The values provided are common starting points and should be measured against the reader's own samples.