Cloud Service Farm Update Orchestration via Tenant Grouping

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Cloud-based services face inconsistencies and security concerns due to staged updates across different regions, leading to users experiencing different versions of the service, which can result in an inconsistent user experience and potential security vulnerabilities.

Innovation Solution

An orchestrator service that manages updates by grouping farms into common stages using a grouping table, ensuring that farms with tenants in different regions are updated simultaneously, thereby maintaining a consistent user experience and enhancing security.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If cloud services are updated in multiple stages across different regions, then the update deployment can be controlled and tested progressively, but different users experience different versions of the service leading to inconsistency and security concerns

Engineering Contradiction:

Improveupdate deployment controlVSAvoidservice version consistency

Core Design Contradiction:

ReliabilityVSStability of the object's composition

Solution Approach 1:

The system segments farms into different update stages (first stage, second stage, third stage) with each stage representing a specific percentage of the userbase (e.g., 10%, 50%, 100%). This segmentation allows controlled progressive deployment while the stage management service ensures that farms serving the same customer are kept in the same stage, preventing version inconsistency within customer organizations.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces a new dimension of customer-tenant relationships overlaying the traditional geographic/regional update stages. By adding the dimension of customer tenancy to the update stage allocation, the system can ensure that all farms serving tenants of the same customer remain in the same update stage, even if those farms are geographically distributed across different regions.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

2Productivity

If updates are deployed to different regions at different times, then regional testing and validation can be performed, but security vulnerabilities arise due to inconsistent service versions across the network

Engineering Contradiction:

Improveregional update testing efficiencyVSAvoidsecurity vulnerabilities

Core Design Contradiction:

ProductivityVSObject-affected harmful factors

Solution Approach 1:

The stage management service continuously monitors and tracks which farms are assigned to which update stages, and uses this feedback information to make real-time adjustments. When a customer has farms in multiple stages, the service detects this inconsistency and reallocates farms to ensure uniform stage assignment, thereby preventing security vulnerabilities from version mismatches.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The stage management service acts as an intermediary between the update deployment process and the farm allocation system. It receives update stage allocation requests, checks for customer tenancy relationships, and mediates the allocation by ensuring that farms serving the same customer are assigned to the same stage, thus preventing security issues while allowing regional testing to proceed.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Adaptability or versatility

If farms are allocated to different update stages based on regional distribution, then update deployment can be managed across global networks, but customers experience inconsistent service versions across their distributed teams

Engineering Contradiction:

Improveglobal network update deploymentVSAvoiduser experience consistency

Core Design Contradiction:

Adaptability or versatilityVSEase of operation

Solution Approach 1:

Before allocating farms to update stages, the stage management service performs preliminary actions by identifying customer-tenant relationships and determining which farms serve which customers. This preliminary identification allows the system to pre-plan stage allocations that keep customer farms together, ensuring consistent user experience across distributed teams while still enabling global network update deployment.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS20240394035A1Updating a cloud service with farms grouped for a same update deployment stage

Publication Date: 2024.11.28 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

An orchestrator for updating a cloud service includes: an orchestrator service host computer comprising a processor and memory; an orchestrator service for execution by the orchestrator service host computer for orchestrating updates to farms of the cloud service over a service bus; a stage management service for determining which farms should be in each stage of an update deployed by the orchestrator service; and a grouping table identifying farms that should be in a common stage of the update deployed by the orchestrator service. The stage management service places farms grouped in the grouping table in a common stage of the update deployment.