Multi-tenant Cloud Update Segmentation for Downtime Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In multi-tenant cloud computing environments, software updates often cause downtime as they simultaneously affect both TEST and PRODUCTION tenants, making it difficult to detect and address failures early, leading to substantial downtime due to incompatibilities and the need for customers or development teams to re-model and re-deploy integration scenarios.
Innovation Solution
A cloud platform provisioning framework that allows tenants to opt for early software updates, automatically installing new updates prior to the scheduled installation time by verifying eligibility and update preferences stored in a provisioning application data store, thereby reducing downtime and enabling early detection of issues on TEST tenants before affecting PRODUCTION tenants.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If software updates are installed simultaneously for all tenants, then the update process is simple and fast, but it causes substantial downtime and makes it difficult to detect failures early
Solution Approach 1:
The system segments the tenant population into different groups (TEST tenants and PRODUCTION tenants) and applies different update schedules to each segment. TEST tenants receive updates early for validation, while PRODUCTION tenants receive updates after successful validation, thereby enabling failure detection in isolated segments without causing widespread downtime.
Solution Approach 2:
The system performs preliminary software updates on TEST tenants before deploying updates to PRODUCTION tenants. This preliminary action allows failures to be detected and resolved in advance, preventing them from affecting production operations and reducing overall downtime.
2Adaptability or versatility
If customers are asked to re-model integration scenarios to fix incompatibilities, then the integration content can be updated, but it causes downtime in message processing
Solution Approach 1:
The system performs integration content updates in advance on TEST tenants, allowing customers to validate and adjust their integration scenarios before production updates. This preliminary action ensures compatibility is established before production operations are impacted, maintaining message processing continuity.
Solution Approach 2:
The system introduces an intermediary validation phase using TEST tenants as a buffer between update deployment and production impact. This intermediary layer allows integration content to be updated and validated without directly disrupting production message processing.
3Reliability
If software updates are delayed to allow testing, then failures can be detected early, but it increases the overall update time and delays productivity improvements
Solution Approach 1:
The system segments the update deployment process into parallel tracks: TEST tenants receive updates immediately for validation, while PRODUCTION tenants receive updates after a predetermined validation period. This segmentation allows failure detection to proceed without delaying the overall update timeline, as both tracks operate concurrently.
Solution Approach 2:
The system implements periodic update deployment with predetermined intervals between TEST and PRODUCTION updates. This periodic action structure allows systematic failure detection while maintaining a predictable and optimized update schedule that minimizes overall deployment time.
Data Source
AI summary
Methods and systems may be associated with a cloud computing environment. A provisioning application data store may contain information associated with services representing a plurality of multi-tenant cloud applications including, for a first SaaS application, indications that a first tenant qualifies for early software updates and has opted to receive early software updates. A cloud platform provisioning framework may determine that a new software update to the first application is available and is scheduled to be installed within a pre-determined time period (e.g., one day or week). The framework may then access the provisioning application data store to verify that the first tenant qualifies for early software updates and has opted to receive early software updates. As a result of the verification, the cloud platform provisioning framework may automatically install the new software update to the first application for the first tenant prior to the scheduled installation.