What data for this category looks like
The platform serves cloud finance operations, FinOps teams, and account managers. It sources data from cloud billing systems, tag data, order systems, and resource monitoring systems. It produces business documents with daily or monthly update frequencies.
Daily outputs include billing details, cost allocation tables, resource utilization reports, and optimization suggestion sheets. Structured fields include resource identifiers, billing cycles, cost amounts, associated tags, order numbers, and more. Data is stored and transferred as tables or exportable files.
What constraints do these characteristics impose during deployment and upgrade?
Since multiple data sources exist and update schedules are fixed, connect access permissions for multiple external systems in advance during deployment to avoid data pull interruptions.
Reserve resource quotas for batch data processing during upgrades to prevent service overload during high-frequency update periods.
Configure sufficient archived storage space to meet storage requirements for multiple types of business documents. Also ensure connection stability between databases and caches to avoid exceptions when querying historical data.
For internal network deployment scenarios, additionally configure security policies for cross-system access to guarantee normal access by internal users, preventing data synchronization failures caused by network policy restrictions.
How to set configurations
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
MONGODB_URI | Point to the read-write split connection string of the cloud billing-exclusive MongoDB cluster | Used to store high-frequency read-write billing details and cost allocation table data. Read-write splitting reduces pressure on single nodes |
REDIS_URL | Configure an internal network Redis instance, set cache period to 3600 seconds | Caches high-frequency queried resource utilization report data to reduce calls to cloud billing systems |
DB_MAX_LINK | Value range 20–30 | Adapts to batch query concurrency for cloud billing data, preventing query failures caused by exceeding connection limits |
STORAGE_S3_ENDPOINT | Point to the endpoint address of the enterprise internal network object storage service | Used to store archived files of historical billing details and optimization suggestion sheets |
FE_DOMAIN | Configure as a dedicated domain name for internal network access | Ensures internal users access the billing page via the specified domain name, avoiding cross-domain access issues |
FILE_TOKEN_KEY | Configure as a 32-bit random string, set token validity period to 86400 seconds | Used to generate temporary access tokens for billing file downloads, ensuring data access security |
Three common mistakes
- Phenomenon: After multi-node deployment, billing data update nodes cannot sync across nodes, leading to inconsistent cost allocation table data viewed by some users. Cause: Connection parameters for
MONGODB_URIandREDIS_URLwere not uniformly configured. Each node uses an independent storage instance, interrupting the data synchronization link. - Phenomenon: After accessing via nginx reverse proxy for internal network deployment, historical billing files cannot be downloaded. Cause: The configurations for
STORAGE_S3_ENDPOINTandFE_DOMAINwere not correctly associated. The domain name of the file download link does not match the proxy domain name. - Phenomenon: After upgrading to FastGPT v4.17.0, generation of cloud billing resource utilization reports fails. Cause: The value of
DB_MAX_LINKwas not adjusted. The default connection number after the upgrade is insufficient to support batch data calculation, leading to database connection timeouts.
How to confirm configurations are correct
- Log in to the FastGPT backend system configuration panel, check the database connection status, and confirm that the connection test results for
MONGODB_URIandREDIS_URLpass. - Initiate a manual synchronization task for historical billing data, check whether cost allocation table data generated by each node is consistent.
- Attempt to access the billing page via the internal network dedicated domain name, download a historical billing file, and confirm that the file can be opened normally.
- Check the system operation logs, confirm that no error messages about
DB_MAX_LINKconnection number exceeding limits appear.
The values given are common starting points and should be measured against the reader's own samples.
Question material comes from public community discussions. Configuration values are common starting points and should be measured against your own samples. Verified on 2026-10-19.