Cloud Server Rightsizing Using Multi-Dimensional Workload Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional cloud server capacity management systems lack a comprehensive framework for leveraging multi-dimensional data to predict capacity needs, failing to collect, process, and utilize usage data across multiple aspects such as CPU, memory, and network traffic, leading to inadequate proactive adjustments.
Innovation Solution
A system that collects and processes multi-dimensional cloud server usage data through outlier removal and seasonality adjustment to generate a multidimensional operating ratio, computes a workload metric, and uses predictive models to anticipate future capacity needs, enabling proactive capacity adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional systems collect and process multi-dimensional usage data, then prediction accuracy improves, but system complexity increases
Solution Approach 1:
The system segments the complex data processing into distinct modules: data collection, outlier removal, seasonality adjustment, and predictive modeling. Each module handles a specific aspect of the multi-dimensional usage data independently, making the overall complex system more manageable and maintainable while preserving prediction accuracy.
Solution Approach 2:
The patent introduces an intermediary processing layer that includes outlier removal and seasonality adjustment functions. These intermediaries clean and prepare the raw multi-dimensional data before it reaches the predictive model, reducing the complexity burden on the model while maintaining high prediction accuracy.
2Reliability
If the system performs comprehensive data cleansing including outlier removal and seasonality adjustment, then data quality improves, but processing time increases
Solution Approach 1:
The system performs outlier removal and seasonality adjustment as preliminary actions before the actual predictive analysis. By pre-processing the data to remove outliers and adjust for seasonal patterns, the subsequent predictive modeling operates on cleaner, more reliable data, improving overall data quality while the processing time is amortized across the entire data pipeline.
3Measurement precision
If the system collects usage data across multiple dimensions, then capacity prediction accuracy improves, but data collection complexity increases
Solution Approach 1:
The data collection system is designed with universality to handle multiple dimensions of usage data (CPU usage, memory usage, disk I/O, network traffic) through a unified collection framework. This multi-functional approach allows the system to gather comprehensive multi-dimensional data without proportionally increasing collection complexity, as the same infrastructure handles all data types.
Data Source
AI summary
In some aspects, systems and methods are described herein for determining rightsizing adjustments to a cluster of cloud servers using a multi-dimensional operating ratio. The system collects usage data from a cloud server, comprising multiple dimensions of cloud computation usage. The system then processes the usage data using a cleansing process to generate processed usage data, wherein the cleansing process comprises outlier removal and seasonality adjustment. The system compares the processed usage data against a benchmark to generate a workload metric. The workload metric corresponds to values in the multiple dimensions of the cloud computation usage, indicating a distance from the expected usage data. Based on the workload metric and using a predictive model, generating expected capacity needs. Based on the expected capacity needs, determining a set of rightsizing changes, wherein the set of rightsizing changes comprises changes to capacities of the cloud server.