SaaS Metadata Compression with Global, Group, and Tenant Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
SaaS systems face challenges with inadequate cache memory due to the linear increase in metadata size as the number of tenants increases, leading to delayed entity object invocation and degraded system performance.
Innovation Solution
A metadata compression method that identifies global and common properties across tenants, partitions them into groups, and creates a new data structure with separate lists for global, group, and tenant-specific properties, reducing the memory footprint of metadata.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If metadata for every tenant's entity objects is stored in cache memory, then the system can quickly access and process entity objects, but the cache memory size required increases linearly with the number of tenants
Solution Approach 1:
The patent segments metadata into three distinct categories: global properties (common to all tenants), group properties (common to groups of tenants), and tenant-specific properties (unique to each tenant). This segmentation allows the system to store only the necessary unique metadata for each tenant in cache memory while sharing common metadata through hierarchical references, thereby reducing the linear growth of cache memory requirements as tenants increase.
Solution Approach 2:
The patent creates a compressed metadata structure that copies only the unique tenant-specific properties into each tenant's metadata object, while global and group properties are referenced from shared parent structures. This copying approach significantly reduces the memory footprint in cache while maintaining fast access to all necessary entity object metadata.
2Adaptability or versatility
If each tenant has unique metadata for entity objects, then the system can support tenant-specific customizations, but the overall metadata size increases with the number of tenants
Solution Approach 1:
The patent divides metadata into hierarchical segments: global properties at the top level (shared by all tenants), group properties at the intermediate level (shared by groups of tenants), and tenant-specific properties at the leaf level (unique to each tenant). This segmentation enables tenant-specific customizations while avoiding duplication of common properties, thus maintaining adaptability without linearly increasing metadata size.
Solution Approach 2:
The patent merges common properties from global and group levels with tenant-specific properties to create complete tenant metadata. By merging shared properties once at higher hierarchical levels and referencing them from multiple tenants, the system supports tenant-specific customizations without replicating common data, thereby controlling overall metadata size.
Data Source
AI summary
Disclosed herein are system, method, and computer program product embodiments for compressing metadata in a Software-as-a-Service (SaaS) system. In embodiment operates by receiving, from a first tenant of a plurality of tenants, a data access request. The first tenant is identified as belonging to a first group of the plurality of groups, wherein the tenants of each group share one or more group properties. The one or more global properties, one or more group properties, and one or more tenant-specific properties are loaded into a data structure in memory. Based on at least a portion of the data structure, data is retrieved from a database correspondent to the data access request. A response to the data access request is provided to the first tenant based on the data retrieved from the database.