Multitenant Server Clustering via Cross-Cluster Session Migration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing clustering features in traditional application server environments need modification to support multi-tenancy in cloud environments, where users or tenants develop and run applications, requiring efficient resource management and session handling across multiple server clusters.
Innovation Solution
A system and method for supporting clustering in a multitenant application server environment, featuring a domain with multiple server clusters and partitions, where an API initiates resource group migration across clusters, invoking a replication manager to determine session storage and supporting additional clustering features like cluster messaging, leasing, singleton management, and domain front-end load balancing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional clustering features are used in application server environments, then clustering functionality is provided, but multi-tenancy support and cloud environment adaptability are insufficient
Solution Approach 1:
The system segments the application server environment into multiple partitions, each representing a tenant. Each partition can be independently configured and managed, allowing multi-tenancy support while maintaining modular complexity. The replication manager further segments session management by tracking primary sessions per partition across clusters.
Solution Approach 2:
The clustering system is designed to serve multiple functions: traditional clustering, multi-tenancy isolation, session replication, and resource migration. The same cluster infrastructure supports both single-tenant and multi-tenant modes, making the system universally applicable without requiring separate configurations for each use case.
2Reliability
If session replication is implemented across clusters, then availability and fault tolerance are improved, but session management complexity increases
Solution Approach 1:
The replication manager acts as an intermediary between clusters, handling session replication automatically. It queries target clusters for runtime topology, determines where to store primary sessions, and manages the replication process without requiring manual intervention. This mediator approach improves reliability while containing complexity within a dedicated management component.
Solution Approach 2:
The system implements session replication by copying session state to multiple clusters. The replication manager creates copies of primary sessions in target clusters based on runtime topology information, enabling failover and high availability. The copying mechanism is automated and topology-driven, reducing manual session management complexity.
3Productivity
If resource group migration across clusters is enabled, then scalability and resource utilization are improved, but migration overhead and operational complexity increase
Solution Approach 1:
The system performs preliminary actions by querying the target cluster for runtime topology information before migration occurs. The replication manager proactively gathers topology data, determines optimal target locations, and prepares migration paths in advance. This preliminary action reduces migration time and enables more efficient resource allocation when actual migration happens.
Data Source
AI summary
In accordance with an embodiment, described herein is a system and method for supporting clustering in a multitenant application server environment. The system includes a domain with a plurality of server clusters, and a plurality of partitions, wherein each cluster is homogenously configured and targeted by one or more partitions. An application programming interface (API) can be used to initiate a resource group across-cluster migration within a partition. The API can invoke a replication manager, which queries a target cluster for a runtime topology for use in determining where to store primary session; migrates the primary session to a determined server instance in the target cluster. Additional clustering features, such as cluster messaging service, leasing service, singleton service management, session replication, clustered JNDI, and use of domain front-end load balancing, can be supported.