Scope of Environment Support
To ensure the stability, security, and performance of the AI Central platform across different cloud environments, we have established unified planning and tiered support for the system runtime environment, including cloud platforms, networks, identity systems, and model service access environments.
The platform currently uses Microsoft Azure as its primary runtime environment and is gradually evaluating compatibility with other cloud platforms.
Environment Support Levels
Standard Support
-
Refers to runtime environments that have completed comprehensive compatibility testing, performance validation, and security review by Medalsoft.
-
Suitable for production environments, with formal production support (SLA) and long-term maintenance assurance.
-
Currently, only Microsoft Azure is classified as a standard support environment.
Experimental Support
-
Refers to runtime environments that meet basic operating requirements but have not yet passed complete compatibility and performance validation by Medalsoft.
-
Can be used for testing, proof of concept (POC), or internal evaluation, but is not recommended for production environments.
-
Risks related to performance, stability, and compatibility must be assessed and validated by users themselves.
-
Examples: AWS, Alibaba Cloud.
Unsupported
-
Environments that do not meet platform deployment requirements or have security and access restrictions.
Standard Supported Environment (Microsoft Azure)
|
Category |
Recommended Configuration |
Description |
|---|---|---|
|
Resource Group Structure |
Single resource group or separated by environment (DEV / UAT / PROD) |
Separation by environment is recommended for easier permission and policy control |
|
Virtual Network (VNet) |
Enable private subnets and service endpoints |
Avoid public network calls and enhance security |
|
App Service |
2-4 cores / 8GB RAM (minimum standard) |
|
|
Database (Azure PostgreSQL / SQL) |
Starting from 4 cores / 16GB RAM |
Adjust according to data volume and number of concurrent users |
|
Storage (Blob Storage) |
Enable geo-redundancy (GRS) |
Improve reliability and backup recovery capability |
|
Identity System (Entra ID) |
Enable MFA and conditional access policies |
Strengthen account security and compliance |
|
AI Service Access |
Use Azure OpenAI Service / Azure Cognitive Services |
Ensure stable call paths and compliant billing |
|
Cache (Azure Cache for Redis) |
Standard or Enterprise edition, with high availability enabled |
Used for session caching and query result caching to improve system response performance |
User Roles and Environment Mapping
|
User Type |
Environment Usedt |
Permission Description |
|---|---|---|
|
Developer |
DEV / UAT environments |
Can deploy, debug, and configure resources; no production access permissions |
|
Tester |
UAT environment |
Can perform functional and performance testing, with read and validation permissions |
|
Operator |
All environments (including PROD) |
Has permissions for system monitoring, deployment, and emergency handling |
|
Business User |
PROD environment |
Limited to formal system usage, with no configuration or modification permissions |