Module Mirroring for Non-Disruptive Storage Cluster Upgrades
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern storage systems like XtremIO face challenges during non-disruptive upgrades (NDUs) as all management processes need to be restarted and reconnected, leading to potential downtime and performance issues due to the sequential and interconnected nature of modules, which can cause bottlenecks and timeouts.
Innovation Solution
The method involves creating a mirrored set of management processes for a storage cluster undergoing an upgrade, interconnecting these with the original set while the system is actively managed, and performing a swift handover between them, using a shared memory interface to facilitate communication and minimize resource access, thereby enabling near-zero downtime.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all management processes are restarted sequentially during an upgrade, then the new software version is deployed, but system downtime and performance degradation occur
Solution Approach 1:
A mirrored set of management processes running the new software version is created and initialized before the actual upgrade handover. This preliminary setup allows the new version to be ready and tested while the original processes continue to handle workloads, enabling a smooth transition without downtime.
Solution Approach 2:
The patent creates a complete mirrored copy of the management processes with the new software version. This copy operates in parallel to the original processes, allowing the system to test and validate the new version before switching over, thus avoiding service interruption during the upgrade.
2Adaptability or versatility
If management processes are restarted and reconnected, then the new software version is implemented, but connection reestablishment causes hiccups and performance issues
Solution Approach 1:
The original management processes continue to handle workloads without interruption while the mirrored processes are being set up and tested. This continuous operation ensures that system productivity is maintained throughout the upgrade process, with no performance hiccups for end users.
Solution Approach 2:
The system dynamically switches between the original and mirrored management processes. The handover mechanism allows flexible transition where the mirrored processes can take over seamlessly once validated, maintaining system adaptability while avoiding performance degradation associated with hard restarts.
3Reliability
If all processes are restarted with new software version, then the upgrade is completed, but sequential restart causes bottlenecks and timeouts
Solution Approach 1:
The upgrade process is segmented into distinct phases: creating the mirrored set, interconnecting components, testing the mirrored processes, and performing the handover. This segmentation breaks down the complex sequential restart into manageable independent steps, reducing coordination complexity and eliminating bottlenecks.
Solution Approach 2:
The mirrored management processes act as an intermediary between the original processes and the final upgraded state. This intermediary layer allows for controlled transition and validation, simplifying the overall upgrade coordination by providing a buffer zone for testing and verification before full commitment to the new version.
Data Source
AI summary
An aspect of module mirroring during an non-disruptive upgrade includes creating a mirrored set of management processes for a storage cluster that is subject to an upgrade a new software version, interconnecting components of the mirrored set and an original set of the management processes while the storage cluster is actively managed by an original set of management processes, and performing a handover between the management processes of the storage cluster.