Command-type Aware VVol Management Command Scheduler
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing VASA provider systems face inefficiencies in processing storage management commands due to unawareness of VASA protocols, leading to sub-optimal multitasking and potential timeouts for higher-priority requests when inundated with commands of varying timeout periods.
Innovation Solution
Implementing a Command-type Aware VVol Management Command Scheduler in conjunction with a VVol Management Command Queue that selectively dequeues storage management commands based on their respective types and timeout periods, ensuring optimal scheduling and prioritization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the VASA provider processes commands in FIFO order without awareness of VASA protocols, then the operations scheduler can manage threads efficiently, but high-priority commands with short timeout periods may timeout even though they should be processed first
Solution Approach 1:
The patent applies dynamics by making the command queue order dynamic rather than static FIFO. The VASA-aware scheduling layer dynamically reorders commands based on their timeout periods and priority levels, allowing short-timeout commands to be processed before long-timeout commands when necessary, thus preventing timeouts while maintaining efficient thread management
Solution Approach 2:
The patent changes the parameter of command scheduling from simple FIFO ordering to priority-based ordering considering timeout periods. By introducing timeout period as a scheduling parameter, the system can adjust command processing order to match urgency levels, ensuring high-priority commands are not delayed by low-priority ones
2Productivity
If the VASA provider is inundated with many commands, then more commands can be processed in parallel, but the operations scheduler performs multitasking sub-optimally because it is unaware of VASA protocols and timeout requirements
Solution Approach 1:
The patent introduces a VASA-aware scheduling layer as an intermediary between the command queue and the operations scheduler. This intermediary translates VASA protocol requirements into scheduling decisions, allowing the operations scheduler to process commands in parallel efficiently without needing to understand VASA protocols itself, thus maintaining productivity while managing complexity
Solution Approach 2:
The patent segments the scheduling function into two parts: the VASA-aware scheduling layer that handles protocol-specific logic and the operations scheduler that handles general thread management. This segmentation allows each component to focus on its strengths, improving overall productivity without overwhelming the operations scheduler with protocol complexity
3Ease of operation
If all long timeout period VASA commands are scheduled to complete before short timeout period VASA commands, then the operations scheduler maintains simple FIFO ordering, but short timeout period requests fail due to timeout even though they are higher priority
Solution Approach 1:
The patent applies preliminary action by sorting and prioritizing commands before they are handed to the operations scheduler. The VASA-aware scheduling layer pre-processes the command queue to ensure high-priority, short-timeout commands are positioned for earlier execution, preventing timeout failures before the actual scheduling and execution occurs
Data Source
AI summary
Embodiments are directed to a method performed by a computing device. The method includes (a) receiving, by the computing device, a stream of storage management commands directed at logical disks hosted by a DSS, the logical disks being accessible to VMs running on a remote host, each storage management command having a command type of a plurality of command types, each command type of the plurality of command types having a respective timeout period, (b) placing the storage management commands of the stream into a VM storage management queue stored on the computing device, and (c) selectively dequeueing storage management commands stored in the VM storage management queue to be performed by the DSS, wherein selectively dequeueing includes applying a set of dequeueing criteria, the set of dequeueing criteria including a criterion that selects storage management commands from the VM storage management queue according to their respective command types.