Hypervisor Upgrade via Virtual Cluster Spanning

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Upgrading blade servers and hypervisors in clustered network environments is a complex, time-consuming, and error-prone process that can disrupt critical applications and pose cross-organizational challenges due to the need to suspend virtual traffic manager instances, leading to costly disruptions.

Innovation Solution

The solution involves upgrading hypervisors in a hardware cluster by spanning virtual clusters across multiple nodes, allowing for continuous operation during upgrades by isolating and upgrading one node at a time, and using pluggable translation modules to manage communication between different versions of hypervisor management control planes and virtual traffic managers.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If hypervisors are upgraded in a clustered network environment, then software version and functionality are improved, but virtual traffic manager instances must be suspended causing service disruption

Engineering Contradiction:

Improvesoftware versionVSAvoidservice continuity

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The system segments the virtual cluster into multiple virtual traffic manager instances distributed across multiple hypervisor nodes. This allows individual nodes to be upgraded while others continue serving traffic, enabling incremental upgrades without complete service suspension.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces translation modules as intermediaries between virtual traffic manager instances and the hypervisor management control plane. These modules enable communication between different software versions, allowing upgraded and non-upgraded instances to coexist and communicate seamlessly during transition periods.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Productivity

If blade servers are upgraded, then hardware performance is improved, but the upgrade process is time-consuming and error-prone

Engineering Contradiction:

Improvehardware performanceVSAvoidupgrade time

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The system performs preliminary actions by creating virtual cluster instances across multiple hypervisor nodes before initiating upgrades. This pre-distribution ensures that service continuity is maintained from the outset, and upgrades can proceed in a controlled, sequential manner rather than requiring complete system shutdown.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent employs parameter changes by modifying the deployment configuration to span virtual clusters across multiple nodes with different software versions. This allows the system to operate with heterogeneous versions during transition, enabling parallel upgrade paths and reducing total upgrade time.

Inventive Principle:

Principle #35Parameter changes

3Adaptability or versatility

If virtual traffic manager instances are suspended for upgrades, then software updates are applied, but critical applications experience costly disruptions

Engineering Contradiction:

Improvesoftware updateVSAvoidapplication disruption

Core Design Contradiction:

Adaptability or versatilityVSObject-affected harmful factors

Solution Approach 1:

The system merges multiple virtual traffic manager instances across different hypervisor nodes into a single virtual cluster that operates as a unified entity. This distribution allows the cluster to maintain service capacity even when individual instances are upgraded, as other instances continue handling traffic.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The patent ensures continuity of useful action by designing the virtual cluster architecture where traffic can be dynamically routed among multiple instances. During upgrades, traffic is continuously handled by non-upgraded instances, and seamlessly transitioned to upgraded instances once they become available, maintaining uninterrupted service.

Inventive Principle:

Principle #20Continuity of useful action

4Productivity

If hypervisors host multiple virtual traffic managers, then resource utilization is improved, but upgrading the hypervisor stops all virtual machine operations

Engineering Contradiction:

Improveresource utilizationVSAvoidvirtual machine operation

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system segments virtual machine operations across multiple hypervisor nodes rather than consolidating them on a single node. This segmentation allows individual hypervisors to be upgraded while other nodes continue operating virtual machines, maintaining overall system availability despite high resource utilization requirements.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS10091138B2In service upgrades for a hypervisor or hardware manager hosting virtual traffic managers

Publication Date: 2018.10.02 F5 NETWORKS INC

AI summary

Embodiments are directed towards upgrading hypervisors operating in hardware clusters that may be hosting one or more virtual clusters of virtual traffic managers. Virtual clusters may be arranged to span multiple computing devices in the hardware cluster. Spanning the virtual clusters across multiple hardware nodes the virtual cluster may enable the virtual clusters to remain operative while one or more hardware nodes may be upgraded. Hypervisor may include a management control plane for virtual clusters of virtual traffic managers. Hypervisors running on hardware nodes may manage the lower level networking traffic topology while the virtual traffic managers may manage the higher level network processing. Further, hypervisor based management control planes may interface with the virtual clusters and virtual traffic manager's using pluggable translation modules may enable different versions of hypervisor based management control planes and virtual traffic managers to communicate and cooperatively manage network traffic.