What this use case’s data looks like
Data for oil and gas exploration and development declaration materials and file review comes primarily from exploration project approval reports, drilling construction logs, reserve assessment documents, environmental assessment approval files, and similar sources. Update frequency changes based on project phase. Phase-based materials are updated quarterly during the exploration phase. After project completion and archiving, the data becomes static for long-term storage. Most documents are long text, with fixed structured fields including well ID, exploration block, approval document number, porosity, permeability, and others. They also include unstructured attachments such as seismic profiles and core photos. Total data volume grows continuously as the project advances.
What constraints do these characteristics impose on deployment and upgrade
The characteristics of oil and gas exploration and development declaration materials and file review data impose clear constraints on the deployment and upgrade link. Long documents and large-volume attachments require database configurations that support high-capacity storage and high-concurrency read/write operations. Mixed structured and unstructured data formats require the system to support unified access and indexing of multi-source data. Continuous data updates during the project cycle require incremental synchronization mechanisms to avoid repeated imports. Long-term archived file data requires that existing review processes are not interrupted during upgrades. Gray release solutions must be used to ensure business continuity.
How to set the configurations
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
DB_MAX_LINK | 200–300 | Single-batch import volume of oil and gas exploration files is large. Sufficient database connection pool is needed to support concurrent review requests across multiple nodes |
STORAGE_S3_ENDPOINT | Dedicated endpoint compatible with enterprise-level S3-compliant storage | Attachments such as seismic profiles and core photos accompanying exploration data have large file sizes. Storage services that meet the security specifications of state-owned enterprises and public sector entities must be used |
REDIS_URL | Dedicated cache instance with memory specification no less than 8 GB | Cache temporary status during review processes and user sessions to avoid slowing overall response speed from frequent database reads |
PG_URL | Independent PostgreSQL instance with reserved storage capacity of at least 500 GB | Stores structured declaration fields and review logs. Long-term archived data volume grows continuously as the project progresses |
FILE_TOKEN_KEY | Complex security key compliant with enterprise information security specifications | Generates access tokens for attachments to ensure access permissions for unstructured attachments meet internal security requirements |
Three common misconfigurations
- Symptom: Workflow canvas drag lag, with database connection pool exhausted errors in backend logs. Cause:
DB_MAX_LINKnot set to a value adapted for high concurrency, leading to insufficient database connections during multi-node reviews. - Symptom: Imported exploration attachments cannot be previewed normally, returning 403 permission errors. Cause:
FILE_TOKEN_KEYnot configured correctly, resulting in access tokens that do not match the verification rules of the enterprise storage service. - Symptom: Knowledge base retrieved file content does not match the query, with large deviations in result counts. Cause: Insufficient reserved database storage capacity, leading to partial historical files not being fully indexed.
How to confirm the configuration is properly set
- Log in to the FastGPT v4.17.0 system backend monitoring panel, check the active connection count of the database connection pool, and confirm it does not exceed the configured
DB_MAX_LINKvalue. - Upload a single exploration attachment compliant with enterprise storage specifications, and verify that it can be previewed and downloaded normally.
- Start a review workflow with multiple nodes, drag the canvas to confirm no significant delay in operation response.
- Import a batch of historical declaration files, and confirm that all preset structured fields can be retrieved normally.
The values provided here are common starting points, and should be evaluated against the user’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 2026-09-28.