HPC Data Protection via Intermediary Decryption
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-performance computing systems face challenges in ensuring secure access control and data protection for batch jobs, as existing encryption techniques lack mechanisms to prevent unauthorized access, provide audit trails, and manage decryption keys effectively, especially when users are not connected.
Innovation Solution
A data protection system with a policy server and data encryption process that restricts output to authorized compute nodes, manages decryption keys securely, and provides audit trails by maintaining policies for access control, encrypting and decrypting data within compute nodes, and ensuring only authorized access to encrypted data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is encrypted and stored at file servers, then data security is improved, but authorized users and batch jobs cannot access the decrypted data when users are not connected
Solution Approach 1:
A data encryption process is introduced as an intermediary component that runs on compute nodes. This intermediary manages decryption keys locally and handles all decryption operations, allowing authorized batch jobs to access encrypted data without requiring user connection. The intermediary mediates between the encrypted data storage and the computational processes needing access.
Solution Approach 2:
The system separates the key management function from user authentication. Instead of requiring users to be connected for key access, the decryption keys are segmented and managed locally by the data encryption process on each compute node. This segmentation allows key access to be independent of user connection status.
2Ease of operation
If decryption keys are provided to applications, then data access is improved, but security control is worsened as applications may leak decrypted data
Solution Approach 1:
The data encryption process serves as an intermediary that holds and manages decryption keys, rather than providing keys directly to applications. This intermediary architecture maintains security control while enabling data access, as the intermediary can audit and control key usage without preventing authorized batch jobs from accessing decrypted data.
Solution Approach 2:
The system implements auditing mechanisms that provide feedback on data access and key usage. The data encryption process tracks and logs all decryption operations, creating an audit trail that enables security monitoring and control while maintaining the ability to access decrypted data when authorized.
3Ease of operation
If batch jobs execute with user access control permissions, then user data access is improved, but unauthorized access to protected data is worsened
Solution Approach 1:
The data encryption process acts as an intermediary authentication layer between batch jobs and encrypted data. It verifies authorization independently of user connection status and enforces access control policies, preventing unauthorized access while maintaining legitimate user data access through the intermediary's authorization checks.
Solution Approach 2:
The system performs preliminary authorization verification through the data encryption process before batch jobs access encrypted data. This preliminary action establishes security controls in advance, ensuring that only authorized batch jobs can access protected data while maintaining ease of operation for legitimate users.
4Quantity of substance
If compute nodes have no local file storage, then resource efficiency is improved, but data access speed is worsened for batch jobs
Solution Approach 1:
The data encryption process on compute nodes serves as a local intermediary that caches decryption keys and manages decrypted data temporarily. This intermediary layer enables fast local access to decrypted data during batch job execution while maintaining the overall architecture where persistent storage remains centralized, preserving resource efficiency.
Data Source
AI summary
A data protection system is provided that allows applications to access protected data in a way that restricts applications from outputting to unauthorized targets any unprotected data derived from the protected data and that ensures that the applications do not have access to a key that allows access to the unprotected data. The data protection system provides a policy server that may execute on a service node of a high performance computing system and a data encryption process that may execute on each compute node that is allocated to an application or batch job. The policy server maintains policies of entities specifying access control for protected data. The data encryption process generates a secure execution environment for an application process and interfaces with the policy server to retrieve keys for decrypting protected data in accordance with a policy, and it decrypts and provides the decrypted data to the application process.