Multi-tenant Hosted Application Scale Group Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Hosted multi-tenant applications face challenges such as simultaneous database outages affecting all tenants, reduced performance due to excessive load, difficulty in customizing database schema, complex and expensive maintenance of separate virtual servers, and limitations in adding hardware capacity or upgrading software without causing downtime.
Innovation Solution
Implementing a hosted application system with a shared middle-tier and a database-tier organized into site-wide servers and scale groups, where each tenant has a private, unshared database, allowing for geographic distribution and load balancing to reduce latency and minimize downtime during upgrades.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If a shared database-tier is used among all tenants, then hardware resource utilization is improved, but system reliability deteriorates because outages affect all tenants simultaneously
Solution Approach 1:
The patent segments the database tier into shared and private components. A shared database tier stores common data accessible by multiple tenants, while each tenant has a private database tier for isolated data storage. This segmentation allows the system to maintain high hardware utilization through sharing while ensuring reliability through isolation - outages or performance issues in one tenant's private database do not affect others.
2Reliability
If separate virtual servers are provided for each tenant, then system reliability is improved by isolating outages, but device complexity and maintenance cost increase
Solution Approach 1:
The patent merges the middle-tier application logic across tenants into a single shared platform, eliminating the need for separate virtual servers for each tenant. Each tenant retains a private database tier for data isolation, but the computational and application layers are consolidated. This merging reduces maintenance complexity and operational overhead while preserving reliability through database-level isolation.
3Use of energy by moving object
If a shared database-tier is used, then hardware resource utilization is improved, but application performance deteriorates when one tenant places excessive load
Solution Approach 1:
The patent segments the database into shared and private portions. The private database tier for each tenant acts as an isolation barrier that prevents excessive load from one tenant from impacting the shared database and other tenants. This allows the system to maintain high hardware utilization through shared resources while preserving application performance through tenant-level isolation.
4Productivity
If separate virtual servers are provided for each tenant, then application performance is maintained, but loss of energy increases due to unused allocated hardware resources
Solution Approach 1:
The patent merges the middle-tier and shared database resources into a single pooled infrastructure that serves multiple tenants. This consolidation eliminates the wasted hardware resources associated with dedicated virtual servers for each tenant, as resources are dynamically allocated and shared based on actual demand. Application performance is maintained through the private database tier isolation mechanism.
Data Source
AI summary
Technologies are described herein for implementing a hosted multi-tenant application system. The server computers utilized to provide the hosted application are organized into logical groupings of server computers called scale groups. One or more tenants are assigned to each scale group. When a new tenant is provisioned, the tenant is assigned to a scale group and a database server in the assigned scale group creates a database for the tenant. An association between the tenant and the scale group is also created in a shared configuration database. When a request is received from a tenant to access the hosted application, the shared configuration database is consulted to locate the scale group hosting the tenant. Once the appropriate scale group has been located, the request is redirected to the appropriate scale group for processing.