Availability Zone Resets in Ultraconverged Servers for Storage Uptime
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Hyperconverged systems face issues with unpredictable storage availability due to shared hardware, leading to potential QoS compromises and prolonged downtime during OS crashes or reboots, which existing Smart NICs do not fully address.
Innovation Solution
Implementing an ultraconverged architecture with multiple availability zones and a manager of availability zones that controls independent resets and reinitializations across zones, using PCIe or other fabrics, ensuring components in one zone can be replaced or rebooted without affecting others.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If hyperconverged architecture shares storage and compute on same hardware, then infrastructure costs are reduced, but storage availability and QoS are compromised
Solution Approach 1:
The system divides the server into multiple availability zones (AZs), separating storage components (Smart NICs and local storage devices) from compute components (main CPU and OS). Each AZ can be independently managed and reset, allowing storage to remain available even when compute resources are rebooted or experiencing issues. This segmentation resolves the contradiction by maintaining physical integration for cost efficiency while achieving logical independence for reliability.
2Productivity
If Smart NICs offload storage services workload, then compute performance is improved, but storage availability during OS crash remains compromised
Solution Approach 1:
The availability zone manager enables independent reset of the Smart NIC availability zone from the main system zone. When the OS crashes or needs rebooting, the Smart NIC zone can be reset independently, allowing storage services to resume quickly without waiting for full system recovery. This maintains the performance benefits of offloading while resolving the availability issue during OS failures.
Solution Approach 2:
The system pre-configures independent reset capabilities for each availability zone through the availability zone manager. This preliminary setup ensures that when failures occur, the storage zone can be rapidly reinitialized without depending on the state of the compute zone, reducing downtime and maintaining storage availability during OS crashes.
3Stability of the object's composition
If OS reboot is required for updates, then system stability is improved, but downtime is prolonged
Solution Approach 1:
By segmenting the system into independent availability zones with separate reset control, the patent allows the storage zone to remain operational or be quickly reinitialized while the compute zone undergoes OS updates. The availability zone manager coordinates resets to minimize overall downtime, resolving the contradiction between maintaining stability through reboots and reducing service interruption time.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An ultraconverged architecture has multiple availability zones within a single server. The functionality in each of the availability zones is independently controlled, such that resetting and/or disconnecting any component in any availability zone from power and replacing said component does not affect availability of any other component in any other availability zone. A manager of availability zones controls reset functionality in each of a plurality of availability zones. The manager of availability zones generates a requested reset type in the requested availability zone. The manager of availability zones generates reset signals or requests for some or all components located in multiple availability zones. The reset signal or request is generated upon external request to the manager of availability zones that specifies the reset type, the availability zone, and optionally the list of components to be reset. The manager of availability zones discovers and enumerates the components in each availability zone.