Deployment and Upgrade for Identity and Timing Insurance

What this type of data looks like

Identity and timing insurance claim initial review data comes from two main sources. First, scanned copies or photos of physical documents submitted by claim applicants, including ID cards, bank cards, and hospital diagnosis certificates. Second, structured data integrated with public security household registration systems, social security payment systems, and claim process backends.

Data update rhythm falls into two categories: single submission and real-time synchronization. Document attachments are uploaded once. Structured timing data updates in real time with claim process nodes. This includes incident report time, application submission time, diagnosis start and end time, and claim time limit requirements.

Document structure uses mostly structured fields, with attached unstructured files. Core fields include identity_card_number, claim_apply_time, hospital_admission_time, and claim_deadline. Time fields use ISO 8601 format uniformly. ID number fields follow the 18-digit number plus check code format.

What constraints do these characteristics impose on deployment and upgrade

The real-time verification and sensitive nature of identity and timing data creates clear constraints for deployment and upgrade.

Integrating third-party identity verification interfaces requires pre-configuring interface keys and whitelists. Network outbound permissions must be reserved during deployment to prevent verification failures caused by firewall blocking.

Structured timing data must synchronize in real time with claim system process nodes. The database must support high-concurrency writes and queries. Sharded clusters must be configured during deployment to handle bulk claim scenarios.

Upgrade processes cannot interrupt real-time verification links. This will stall initial review workflows. Blue-green deployment or rolling upgrade strategies must be used. Configuration rules for different regions must also be compatible, to avoid adaptation failures after upgrades caused by hard-coded rules.

How to set configurations

Configuration Item	Recommended Value	Rationale
`MONGO_CONNECTION_STRING`	`mongodb://fastgpt:password@localhost:27017/fastgpt_claim?authSource=admin`	Identity and timing claim data requires stable structured storage. This parameter specifies the MongoDB connection address, and must match the deployment environment's instance configuration
`THIRD_PARTY_API_TIMEOUT`	`10-15 seconds`	Identity verification integrates with public security interfaces. Excessive timeout will affect initial review efficiency. This parameter controls the wait duration for third-party interfaces
`UPLOAD_FILE_MAX_SIZE`	`10 MB`	Identity-related attachments are mostly ID photos and scanned diagnosis certificates. This parameter limits upload volume to avoid storage and transmission pressure
`PARSE_FILE_TIMEOUT_SECONDS`	`300 seconds`	Parsing long-text claim timing description documents takes significant time. This parameter sets the timeout threshold for file parsing
`RAG_RECALL_TOP_K`	`Top 3-5 entries`	Association rules for identity and timing are mostly fixed items. Excessive recall increases inference load
`ENABLE_HTTPS`	`Yes`	Identity data falls under sensitive information. HTTPS protocol must be used for encrypted transmission to ensure data security

The parameter values provided on this page are common recommended starting points for configuration. Actual values are affected by material form, data volume, and business rules. Specific issues require targeted analysis. It is recommended to test on your own samples before finalizing settings.

Three common mistakes

Issue: pnpm dev displays MongoDB connection timeout during startup, with console error ECONNREFUSED. Cause: The username, password, or database address in MONGO_CONNECTION_STRING is not configured correctly, resulting in failure to connect to local or remote MongoDB instances.
Issue: The system returns a 504 Gateway Timeout error during bulk claim initial review processing. Cause: The THIRD_PARTY_API_TIMEOUT parameter is not adjusted, and the wait duration for third-party identity verification interfaces is insufficient, causing requests to be truncated by the gateway.
Issue: A file size limit exceeded prompt triggers when uploading an ID photo, even though the actual file is smaller than 10 MB. Cause: UPLOAD_FILE_MAX_SIZE is incorrectly configured to a value smaller than the actual file, and does not match the common volume range of ID photos.

How to confirm configurations are correct

Run docker logs fastgpt to view container logs, and confirm there are no error messages about failed MongoDB connections.
Upload an ID photo with a volume of 12 MB, check if a file size limit exceeded prompt triggers, to verify that the UPLOAD_FILE_MAX_SIZE configuration takes effect.
Call the third-party identity verification interface, confirm that the request returns a result within 15 seconds with no timeout errors, to verify that the THIRD_PARTY_API_TIMEOUT configuration is reasonable.
Access the deployment address, check if the address prefix is https://, to confirm that the ENABLE_HTTPS configuration takes effect.

Question material comes from public community discussions. Configuration values are common starting points and should be measured against your own samples. Verified on 2026-09-14.