Server Replication and Learning for Security
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing computer and network security systems, despite using firewalls, lack the ability to evolve and learn from attacks, leading to inefficiencies in resource consumption and protection against corrupt data and queries.
Innovation Solution
A computer-based security system that creates multiple copies of a lead server, allowing seamless query processing and automatic replacement of the lead server when issues arise, with a learning server analyzing problematic queries to generate rules for future prevention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If firewalls and security software are installed on the server side to prevent harmful content, then security protection is improved, but resource consumption increases and the system lacks the ability to evolve and learn from attacks
Solution Approach 1:
The system divides security functions into two segments: a learning server that analyzes attacks and generates security rules, and lead servers that execute queries against those rules. This segmentation allows the learning server to evolve security knowledge without consuming resources of the lead servers during normal operation, resolving the contradiction between security improvement and resource consumption
Solution Approach 2:
The learning server performs preliminary analysis of attacks and generates security rules in advance, before these rules are needed by the lead servers. This preliminary action allows the system to prepare security defenses without impacting the resource usage of servers during normal query processing, enabling evolution and learning without increasing operational resource consumption
2Reliability
If multiple copies of the lead server are created for query processing, then system availability and continuity are improved, but device complexity increases
Solution Approach 1:
The system creates multiple copies of the lead server that can process queries independently. When a problem is detected, one copy can be replaced with a fresh copy, ensuring continuous availability. The copying principle simplifies the complexity management because copies are identical replicas that can be rapidly deployed and replaced without complex reconfiguration
Solution Approach 2:
When a lead server copy becomes compromised or malfunctioning, the system discards that copy and recovers by deploying a fresh copy from the learning server. This discard and recover mechanism maintains system availability while managing complexity through automated replacement rather than manual intervention, reducing operational complexity despite having multiple servers
3Reliability
If the lead server is replaced when problems occur, then system integrity is maintained, but processing time is lost during replacement
Solution Approach 1:
The learning server continuously generates and prepares replacement lead server copies in advance, before they are needed. When a problem occurs, a pre-prepared copy is immediately deployed, minimizing downtime. This preliminary preparation of replacement servers ensures system integrity is maintained while reducing the time loss associated with replacement operations
Data Source
AI summary
In an embodiment, a computer-based service security system receives a communication such as a query. The system processes that query in a server, which may be referred to as a lead server, and creates a plurality of copies of that lead server. The query is transferred to the plurality of copies of the lead server when the query did not cause a problem in the lead server. The query is transferred to a learning server when the query caused a problem in the lead server. The lead server is replaced by one of the copies of the lead server when the query caused a problem on the lead server.