Common Key Routing for Telecommunication Data Locality
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Stateful application servers in telecommunications networks, such as policy and charging application servers, face challenges in scaling due to their requirement for maintaining session stores and high-availability redundancy, which limits their ability to utilize existing scalable computing resources effectively.
Innovation Solution
A method and system that utilize a multiprocessor computing system with a router processor to manage communication messages by generating a common key based on subscriber identifiers, identifying logical scalable units, and routing messages to appropriate application processors, while updating data locality tables to adapt to scaling operations, thereby enabling stateful application servers to scale elastically and maintain data locality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If stateful application servers maintain session stores for high availability, then reliability is improved, but scalability deteriorates
Solution Approach 1:
The system segments stateful session stores into distributed components across multiple logical scalable units. Each unit maintains its own session store, allowing independent scaling and failure isolation. The router processor routes messages to appropriate units based on subscriber identifiers, enabling both high availability through redundancy and scalability through distributed architecture.
Solution Approach 2:
The router processor acts as an intermediary between message sources and application servers. It maintains data locality tables that map subscriber identifiers to specific logical scalable units, enabling intelligent routing. This intermediary layer decouples the scaling complexity from application servers, allowing them to scale independently while maintaining session state and high availability.
2Adaptability or versatility
If data is stored remotely from application processors, then scalability is improved, but latency increases
Solution Approach 1:
The system implements local quality by storing subscriber data in session stores co-located with the logical scalable units that process messages for those subscribers. The router processor routes messages to the specific logical scalable unit where the subscriber's session data is stored, ensuring data is accessed locally rather than remotely. This eliminates network latency for data access while maintaining scalability through the distributed architecture.
3Reliability
If common key routing is used to route messages, then data locality is improved, but routing complexity increases
Solution Approach 1:
The router processor performs preliminary action by pre-computing and storing data locality tables that map subscriber identifiers to logical scalable units before message processing occurs. These tables are updated when subscribers are routed to different units, enabling fast lookup during message routing. This preliminary computation eliminates the need for complex real-time routing decisions, reducing operational complexity while maintaining excellent data locality.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A combination of a common key routing (CKR) method and a data locality table (DLT) method may be used to route communication messages in a telecommunication network to achieve improved data locality and reduce secondary key lookups. A processor in a multiprocessor computing system may receive a first communication message that includes subscriber identifiers that uniquely identify a subscriber, generate a common key based on the subscriber identifiers included in the first communication message, and identify a first logical scalable unit that includes a first database memory that stores subscriber data relating to the subscriber by querying a data locality table with the common key. The processor may also identify a first application processor in the multiprocessor computing system that is part of the first logical scalable unit and send the first communication message to the first application processor.