Blade Cluster Switching Server Resource Pooling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing switching center server architectures with a blade cluster structure face challenges in efficiently managing TDM terminations and BICC call instance codes, leading to complex configurations and resource partitioning issues, which hinder scalability and resilience.
Innovation Solution
A switching center server with a distributor and a master that centrally controls pooled resources, allowing messages to be routed to any blade based on processing load and resource availability, providing a single node view and enabling seamless addition or removal of blades without affecting other nodes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If TDM terminations and CICs are partitioned and administratively assigned to particular MSC blades, then resource management is simplified for each blade, but the configuration complexity increases and scalability is hindered
Solution Approach 1:
The patent segments the blade cluster into multiple groups, where each group shares a common pool of TDM terminations and CICs. This segmentation allows independent management of resource pools while maintaining scalability, as new blades can be added to existing groups or create new groups without reconfiguring the entire system.
Solution Approach 2:
The patent creates universal resource pools that can be shared across multiple blade groups. TDM terminations and CICs are no longer dedicated to single blades but become universal resources that any blade in the group can access, enabling flexible resource allocation and improving scalability.
2Reliability
If one blade fails in a partitioned system, then the failed blade is isolated, but the resources assigned to it become unavailable for call connections
Solution Approach 1:
The patent merges resource management at the group level rather than blade level. Multiple blades share common pools of TDM terminations and CICs, so when one blade fails, its resources are automatically available to other blades in the same group, maintaining both reliability and productivity.
Solution Approach 2:
The patent implements local resource sharing within blade groups. Each group has its own pooled resources that can be locally reallocated upon blade failure, providing quick recovery without affecting other groups. This local quality approach maintains high availability while preserving system modularity.
3Productivity
If the number of blades is changed to increase processing power, then system capacity is improved, but the media gateway and resource partitioning must be re-configured
Solution Approach 1:
The patent implements dynamic resource pooling where the association between blades and resources is flexible rather than static. When blades are added or removed, the system dynamically reallocates resources within groups without requiring manual reconfiguration of media gateways or resource partitions, enabling easy scaling.
Solution Approach 2:
The patent introduces blade groups as an intermediary layer between individual blades and resource pools. This intermediary abstracts the complexity of resource management, allowing blades to be added or removed without direct impact on media gateway configuration. The group level handles resource allocation dynamically.
4Productivity
If TDM terminations are controlled by a blade cluster system, then system capacity is increased, but inter-blade coordination is required for termination seizure
Solution Approach 1:
The patent extracts the coordination function from individual blade interactions and consolidates it at the blade group level. The group-level resource pool manages termination seizure coordination, eliminating the need for complex peer-to-peer coordination between multiple blades while maintaining increased system capacity.
Data Source
AI summary
The present invention relates to a switching center server (100) handling calls, comprising a blade cluster with a plurality of different blades (110), a distributor (125, 170) distributing a message relating to a call to one of the blades, a master (130, 150) centrally controlling a usage of pooled resources by said plurality of blades (110) for handling the call, wherein the distributor contacts the master for retrieving the information to which blade said message is to be distributed. The different blades in the blade cluster are seen from the outside as one node, so that for message distribution the master is contacted in order to distribute the message relating to a call to the blade that is handling said call.


