Dynamic Workload Throttling for Throughput SLO Stability

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current throttling mechanisms in data storage networks are ineffective and result in significant oscillations, leading to suboptimal throughput and failure to meet guaranteed minimum throughput service level objectives (SLOs) for high-priority workloads.

Innovation Solution

A dynamic throttling method is implemented by a throughput management module that monitors performance intervals, sets incremental or full throttles based on detection intervals, and adjusts throttle rates to optimize throughput, reducing oscillations and ensuring compliance with SLOs.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If current throttling mechanisms are used to meet minimum throughput SLOs, then priority workload requirements may be satisfied, but significant oscillations occur resulting in too much or too little throttling and suboptimal throughput

Engineering Contradiction:
ImproveSLO satisfactionVSAvoidthroughput
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent implements dynamic throttling that continuously adjusts the throttle rate based on real-time monitoring of priority workload throughput and SLO compliance. The system transitions from static throttling to adaptive throttling where the throttle rate is dynamically modified according to current system conditions, eliminating oscillations and achieving optimal throughput while reliably meeting SLOs

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system employs feedback mechanisms by monitoring priority workload throughput and SLO compliance in real-time, then using this information to adjust the throttle rate applied to non-priority workloads. This closed-loop control ensures that throttling actions are based on actual system state, preventing both over-throttling and under-throttling while maintaining optimal performance

Inventive Principle:
Principle #23Feedback

2Productivity

If resource utilization is increased to improve throughput, then more workloads can be handled, but latency increases and the guaranteed minimum throughput SLO fails to be satisfied

Engineering Contradiction:
ImprovethroughputVSAvoidlatency
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent changes the throttle rate parameter dynamically based on system conditions. By adjusting this control parameter in response to monitored performance metrics, the system optimizes the balance between throughput and latency, ensuring SLO compliance while maximizing resource utilization efficiency

Inventive Principle:
Principle #35Parameter changes

3Reliability

If throttling is applied to non-priority workloads to meet SLOs, then priority workload throughput may be guaranteed, but the throttle may not be effective and unnecessary throttling occurs

Engineering Contradiction:
Improvethroughput guaranteeVSAvoidoverall throughput
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The system uses feedback from SLO compliance monitoring to determine whether throttling is actually needed. By continuously checking if priority workload throughput meets the guaranteed minimum, the system applies throttling only when necessary, avoiding unnecessary performance degradation while ensuring SLOs are met when required

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP4004723B1Methods for dynamic throttling to satisfy minimum throughput service level objectives and devices thereof
Publication Date: 2025.12.17 NETAPP INC
  • EP4004723B1 patent drawingFigure 1
  • EP4004723B1 patent drawingFigure 2
  • EP4004723B1 patent drawingFigure 3

AI summary

Methods, non-transitory machine readable media, and computing devices that dynamically throttle non-priority workloads to satisfy minimum throughput service level objectives (SLOs) are disclosed. With this technology, a determination is made when a number of detection intervals with a violation within a detection window exceeds a threshold, when a current one of the detection intervals is outside an observation area. The detection intervals are identified a violated based on an average throughput for priority workloads within the detection intervals exceeding a minimum throughput SLO. A throttle is then set to rate-limit non-priority workloads, when the number of violated detection intervals within the detection window exceeds the threshold. Advantageously, throughput for priority workloads is more effectively managed and utilized with this technology such that throttling oscillations are reduced, throttling is not deployed in conditions in which it would not improve throughput, and throttling is minimally deployed to maximize throughput.