Kernel Module Database Process Categorization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In computing systems, administrators face challenges in correlating processes with specific database instances due to obscure process identifiers and the lack of native OS implementations for tracking higher-level application processes, leading to difficulties in managing resource allocation and addressing 'noisy neighbor' issues where one database instance starves another of resources.
Innovation Solution
A system and method for kernel-level monitoring of process-based OS resource utilization by individual database instances in a multi-database environment, using a virtual filesystem and kernel-based communications modules to track and categorize processes associated with database instances, avoiding the need for agent-based tracking and providing accurate, real-time resource utilization data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional OS process identification methods are used, then process tracking is implemented, but process correlation to database instances becomes extremely difficult due to obscure identifiers
Solution Approach 1:
The patent introduces a kernel module as an intermediary between the OS process management system and database instance tracking. This kernel module intercepts process creation events and establishes correlations between processes and database instances at the kernel level, providing meaningful process identifiers that can be easily correlated without requiring administrators to manually interpret obscure process IDs
Solution Approach 2:
The patent replaces the manual/mechanical process of correlating processes to database instances (which requires administrators to manually interpret process listings and match them to database processes) with an automated kernel-level tracking system that automatically establishes and maintains these correlations through system calls and event interception
2Measurement precision
If no kernel-level monitoring is implemented, then system simplicity is maintained, but accurate real-time resource utilization tracking by database instances cannot be achieved
Solution Approach 1:
The kernel module leverages existing OS mechanisms such as system calls, process events, and resource allocation interfaces that are already present in the operating system. By using these existing self-service capabilities, the patent achieves accurate resource tracking without requiring a complete overhaul of the OS architecture or adding excessive complexity
Solution Approach 2:
The kernel module is designed to work with multiple database instances and various resource types (CPU, memory, I/O) through a unified interface. This multi-functional approach allows the same kernel infrastructure to track resources for different database instances simultaneously, reducing overall system complexity compared to having separate monitoring mechanisms for each instance
3Productivity
If agent-based tracking is used, then process monitoring is implemented, but system performance overhead increases and real-time accuracy decreases
Solution Approach 1:
The patent replaces agent-based tracking (which requires separate software agents running in user space to collect and report process information) with direct kernel-level monitoring. The kernel module intercepts system calls and process events at the core OS level, eliminating the need for intermediary agents and reducing the performance overhead associated with agent-based collection and reporting mechanisms
Data Source
AI summary
Described is a system, method, and computer program product to perform monitoring for process-based OS resource utilization by individual database instances in a multi-database environment. This approach may be used to resolve numerous resource allocation and monitoring problems, such as the noisy neighbor problem.