Multi-tenant Metadata Segregation via Shared Global Repository
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional multi-tenant architectures lack support for sharing common metadata across tenants while maintaining tenant isolation, leading to issues such as data loss during upgrades, lifecycle and security complexities, increased storage costs, and inefficient caching.
Innovation Solution
A method for segregating and accessing global and tenant metadata in a multi-tenant environment by determining the namespace of documents to identify whether they are global or tenant-specific, allowing for centralized storage and customization of global metadata without duplication, using a metadata engine and Meta Data Services (MDS) to manage and store metadata.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If common metadata is duplicated into each tenant's storage during tenant provisioning, then tenant isolation is maintained, but storage costs increase and updates to shared metadata are lost during application upgrades
Solution Approach 1:
The patent segments metadata into two distinct categories: global metadata stored in a shared repository accessible by all tenants, and tenant-specific metadata stored in isolated tenant repositories. This segmentation allows the system to maintain tenant isolation for private data while sharing common metadata centrally, eliminating the need to duplicate global metadata across all tenants and thereby reducing storage costs.
Solution Approach 2:
The patent introduces a metadata engine as an intermediary layer between tenants and the metadata repositories. This metadata engine manages access to both global and tenant-specific metadata, enforcing isolation policies while enabling efficient sharing of common metadata. The metadata engine acts as a mediator that allows tenants to access global metadata without direct access to the shared repository, maintaining security and isolation.
2Reliability
If common metadata is duplicated into each tenant's storage, then tenant isolation is maintained, but any updates to shared global metadata will be lost when the application is upgraded
Solution Approach 1:
By segmenting metadata into global and tenant-specific portions, the patent ensures that global metadata is stored once in a shared repository and can be updated centrally without affecting tenant isolation. When application upgrades occur, the global metadata is updated in the shared repository, and all tenants automatically access the updated version, preventing any loss of metadata updates while maintaining tenant isolation.
Solution Approach 2:
The global metadata repository serves as a universal storage location that all tenants access for common metadata. This universal approach allows a single source of truth for global metadata, ensuring that updates made to global metadata are immediately available to all tenants without duplication or loss, while tenant-specific metadata remains isolated in separate repositories.
3Reliability
If common metadata is duplicated for each tenant, then tenant isolation is maintained, but lifecycle and security are co-mingled making patching and upgrading difficult
Solution Approach 1:
The patent segments the metadata management system into distinct global and tenant-specific components with separate lifecycle management. Global metadata in the shared repository can be updated, patched, and upgraded independently of tenant-specific metadata. This segmentation allows the system administrator to manage global metadata lifecycle without affecting individual tenants, simplifying patching and upgrading operations while maintaining tenant isolation.
Solution Approach 2:
The metadata engine serves as an intermediary that manages the complexity of lifecycle operations. It handles patching and upgrading of global metadata centrally, applying updates to the shared repository without requiring changes to individual tenant repositories. This mediator approach simplifies lifecycle management by abstracting the complexity away from tenant-specific operations while maintaining security and isolation boundaries.
4Reliability
If common metadata is duplicated into each tenant's storage, then tenant isolation is maintained, but shared objects cannot be efficiently cached
Solution Approach 1:
By segmenting metadata into global and tenant-specific portions, the patent enables efficient caching of global metadata in the shared repository. Since global metadata is stored once and shared across all tenants, it can be cached at the global level and served to multiple tenants simultaneously, improving caching efficiency and reducing redundant storage while maintaining tenant isolation for private metadata.
Data Source
AI summary
The present disclosure relates generally to multi-tenant environments, and more particularly, to techniques for segregating and accessing (e.g., reading and writing) global and tenant base and customization metadata in a multi-tenant environment. One of the techniques includes determining whether a metadata document to be accessed is a global document, if so, accessing the metadata document from a global repository shared amongst tenants of a multi-tenant system, and accessing the metadata document from a tenant repository that is accessible by a tenant of the multi-tenant system when the metadata document is not a global document. The technique further includes disallowing tenants from directly modifying global metadata to not affect other tenants, but still allow tenants to customize global metadata and store the changes as tenant customization documents in the tenant repository. The tenant repository which can be a database or file system is isolated from other tenants of the multi-tenant system.