Write Pacing Simulation for Storage System Latency and Cache Overflow

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current data replication methods in storage systems face latency issues due to synchronous data transfer modes and lack of data consistency between primary and secondary storage devices, especially when there is a long distance between them, and can result in cache overflow during burst writes, leading to potential data inconsistencies and system slowdowns.

Innovation Solution

A method for simulating write pacing in data storage systems, which includes receiving input on configuration, workload, and events such as snapshot copy operations, to determine processing characteristics and display results, allowing for analysis of potential slowdowns and optimizing write pacing operations to prevent cache overflow and maintain data consistency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If synchronous data transfer mode is used for replication, then data consistency between primary and secondary storage devices is ensured, but latency is significantly increased due to waiting for acknowledgment

Engineering Contradiction:
Improvedata consistencyVSAvoidlatency
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system dynamically switches between synchronous and asynchronous replication modes based on system conditions. The replication mode can be changed from synchronous to asynchronous when latency is a concern, and back to synchronous when data consistency is critical, allowing the system to adapt to varying operational requirements

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes the replication parameter (synchronous/asynchronous mode) to resolve the contradiction. By modifying this key parameter, the system can adjust the balance between data consistency and latency performance based on current operational needs

Inventive Principle:
Principle #35Parameter changes

2Loss of time

If asynchronous data transfer mode is used for replication, then latency is reduced, but data consistency between primary and secondary storage devices cannot be guaranteed

Engineering Contradiction:
ImprovelatencyVSAvoiddata consistency
Core Design Contradiction:
Loss of timeVSReliability

Solution Approach 1:

The system dynamically switches between synchronous and asynchronous replication modes based on system conditions. The replication mode can be changed from synchronous to asynchronous when latency is a concern, and back to synchronous when data consistency is critical, allowing the system to adapt to varying operational requirements

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system uses feedback mechanisms to monitor replication status and system conditions, enabling dynamic adjustment of the replication mode. This feedback loop allows the system to maintain data consistency when needed while allowing latency reduction when appropriate

Inventive Principle:
Principle #23Feedback

3Reliability

If write pacing is applied to prevent cache overflow during burst writes, then cache overflow is prevented, but system performance is reduced due to controlled write rates

Engineering Contradiction:
Improvecache overflow preventionVSAvoidwrite throughput
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The write pacing mechanism dynamically adjusts write rates based on system conditions. When the cache is approaching capacity, the system slows down writes to prevent overflow. When the cache has space, the system increases write throughput, creating a dynamic balance between protection and performance

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system applies preliminary anti-action by proactively slowing down writes before the cache overflows. The write pacing mechanism anticipates potential overflow conditions and preemptively reduces write rates to prevent the harmful effect of cache overflow, rather than reacting after overflow occurs

Inventive Principle:
Principle #9Preliminary anti-action

4Productivity

If simulation tool is used to analyze and optimize write pacing operations, then write pacing occurrences are minimized, but system complexity is increased

Engineering Contradiction:
Improvewrite pacing optimizationVSAvoidsystem complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The simulation tool creates a virtual copy or model of the storage system's write pacing behavior. By simulating the system's response to various conditions, the tool allows analysis and optimization of write pacing strategies without modifying the actual production system, enabling proactive optimization

Inventive Principle:
Principle #26Copying

Data Source

PatentUS8843714B1Write pacing simulation tool
Publication Date: 2014.09.23 EMC IP HLDG CO LLC
  • US8843714B1 patent drawing
  • US8843714B1 patent drawing
  • US8843714B1 patent drawing

AI summary

A tool and/or other system is provided that simulates write pacing in connection with a requested operation for a storage system to allow an analysis of the effect of the requested operation on the storage system. In an embodiment, the storage system may be an RDF system operating in an asynchronous transfer mode (e.g., an SRDF/A system) and the requested operation may correspond to a snapshot copy operation of the R2 device and/or a slowdown in one of the storage devices and/or the links therebetween while the SRDF/A system is operating. The system described herein enables simulation of a storage system to take into account an effect of write pacing and may be advantageously used to anticipate and/or otherwise minimize or avoid operations at a particular time that would create undesired initiation of write pacing.