Multi-Layer Cloud Load Balancing With Selective Session Failover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fault tolerance solutions for stateful services in 5G RAN applications, particularly in multi-layer cloud systems, fail to provide adequate redundancy and failover mechanisms, especially when interactions occur across multiple layers, leading to service disruptions and increased costs due to re-execution of procedures.
Innovation Solution
A method where load balancers in each layer maintain information about main and backup serving entities across adjacent layers, enabling seamless failover by replicating session data within limited scope and minimizing time delays, ensuring fault tolerance in both forward and backward directions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If session data is replicated across all entities in a cluster, then fault tolerance is improved, but device complexity and resource overhead increase
Solution Approach 1:
The patent implements selective session data replication where only specific entities are designated as backup entities for particular main entities, rather than universal replication across all entities. This localizes the fault tolerance mechanism to where it is most needed while reducing unnecessary complexity elsewhere in the system.
Solution Approach 2:
The patent segments the cluster into main entities and backup entities with specific pairing relationships. This segmentation allows fault tolerance to be implemented in a structured, manageable way rather than as a monolithic system, reducing overall complexity while maintaining reliability.
2Reliability
If session data is saved to persistent storage, then fault tolerance is improved, but access time and system response speed decrease
Solution Approach 1:
The patent creates in-memory copies of session data at backup entities rather than accessing persistent storage during failover. This copying mechanism allows rapid retrieval of session data from memory at the backup entity, avoiding the time penalty of disk I/O while ensuring fault tolerance through data duplication.
Solution Approach 2:
The patent performs session data replication in advance from main entities to designated backup entities before failures occur. This preliminary action ensures that session data is already available in memory at backup entities, eliminating the need for time-consuming data retrieval operations during actual failover events.
3Stability of the object's composition
If sticky sessions are used to maintain session state, then service continuity is improved, but fault tolerance deteriorates
Solution Approach 1:
The patent introduces backup entities as intermediaries between main entities and clients. These backup entities maintain copies of session data and can step in when main entities fail, mediating the transition from sticky session constraints to fault-tolerant operation without disrupting session continuity for the client.
Solution Approach 2:
The patent prepares backup entities with replicated session data in advance, creating a cushion of redundancy before failures occur. This beforehand cushioning ensures that when a main entity fails, the backup entity can immediately take over without session loss, combining the benefits of sticky sessions with fault tolerance.
4Reliability
If session data is replicated to backup entities, then failover capability is improved, but memory usage and resource consumption increase
Solution Approach 1:
The patent applies selective replication where only specific backup entities store copies of specific main entity's session data, rather than all entities replicating all session data. This localized approach reduces total memory consumption while maintaining failover capability for critical sessions.
Solution Approach 2:
The patent implements partial replication of session data to backup entities, replicating only the necessary portion required for failover rather than complete duplication of all session state. This partial action reduces memory overhead while providing sufficient redundancy for fault tolerance.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A fault tolerance solution for a multi-layer cloud system. The embodiments of this disclosure are applicable to RAN application area as well as cloud environment with virtual machine or container-based application function implementation. In one aspect there is a method performed by a load balancing system, which method includes: receiving a request sent by a client and selecting from a first cluster of serving entities a first main serving entity and a first backup serving entity. The method also includes, as a result of receiving the request sent by the client, sending to the client information identifying the selected first main serving entity and the selected first backup serving entity.