Cloud Management Server Load Balancing via Hierarchical CI Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Balancing management duties in hybrid cloud systems leads to duplication of management data, increased synchronization overhead, and lower stability due to poor load balancing across servers, resulting from geographic and departmental divisions.
Innovation Solution
The method involves determining relationships between configuration items in a cloud system, grouping them into cluster levels based on proximity, and assigning these levels to management cloud servers based on their capacity to distribute and balance management events effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If management duties are divided by geographic area and department borders, then management can be distributed across multiple servers, but duplication of management data occurs and synchronization overhead increases
Solution Approach 1:
The patent segments the cloud system into hierarchical clusters (first-level clusters containing configuration items, second-level clusters containing first-level clusters). This segmentation allows management duties to be distributed across multiple management cloud servers while avoiding data duplication by ensuring each cluster level is assigned to a specific server based on relationship proximity.
Solution Approach 2:
The patent introduces a hierarchical dimension to cluster organization, creating multiple levels (first-level and second-level clusters) rather than a single flat division. This dimensional approach enables efficient distribution of management duties across servers while maintaining data integrity by organizing clusters based on relationship proximity rather than geographic or departmental boundaries.
2Productivity
If management duties are divided by geographic area and department borders, then distribution across servers is achieved, but load balancing becomes poor and stability decreases
Solution Approach 1:
The patent applies local quality by assigning specific cluster levels to specific management cloud servers based on their capacity and the relationships within those clusters. Each server manages a localized portion of the hierarchy (specific cluster levels) rather than arbitrary geographic divisions, ensuring that servers handle workloads appropriate to their capacity and that related configuration items are managed together, improving load balancing and system stability.
3Adaptability or versatility
If configuration items are grouped by geographic and departmental borders, then management can be decentralized, but synchronization overhead increases between servers
Solution Approach 1:
The patent performs preliminary action by pre-calculating and establishing the hierarchical cluster structure based on relationships between configuration items before management duties are assigned. Configuration items are first grouped into first-level clusters based on their relationships, then these clusters are organized into second-level clusters. This preliminary organization ensures that related items are grouped together, minimizing the need for cross-server synchronization and reducing time losses.
Data Source
AI summary
A method for balancing management duties in a cloud system can include determining a relationship between a number of configuration items (CI) in a cloud system. The number of CIs are grouped into a number of cluster levels based on the relationship of the number of CIs. To balance management duties in the cloud system, one of the number of cluster levels is assigned to a management cloud server based on a capacity of the management cloud server.