Dynamic Checkpoint Interval Adjustment for Network Throughput

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing fault-tolerant computing systems face performance degradation due to excessive checkpointing, particularly under streaming network loads, which can lead to decreased network throughput and compromised system utility.

Innovation Solution

A dynamic method for determining the checkpoint interval delay based on network parameters, such as target bit transfer rate and batch size, is implemented to reduce the checkpoint rate proportionally, allowing for flexible and responsive checkpointing without penalizing latency-sensitive loads.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If periodic checkpointing is implemented to maintain synchrony between virtual machines, then fault tolerance is improved, but network throughput deteriorates due to excessive checkpointing overhead

Engineering Contradiction:
Improvefault toleranceVSAvoidnetwork throughput
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The checkpointing system transitions from static periodic checkpointing to dynamic event-driven checkpointing. The checkpoint interval is no longer fixed but adapts based on network events and data changes. This allows the system to increase checkpoint frequency when needed for fault tolerance while reducing frequency during stable periods to maintain network throughput.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The checkpointing parameters (interval, frequency, timing) are changed dynamically based on system conditions. Instead of using a fixed periodic interval, the system adjusts checkpoint parameters according to network activity, data change rates, and synchronization requirements, optimizing both reliability and productivity.

Inventive Principle:
Principle #35Parameter changes

2Reliability

If frequent checkpointing is performed to ensure data synchrony, then data loss prevention is improved, but system performance deteriorates due to processing overhead

Engineering Contradiction:
Improvedata loss preventionVSAvoidsystem performance
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The checkpointing system incorporates feedback mechanisms that monitor data changes, network events, and synchronization status. Based on this feedback, the system intelligently determines when checkpointing is necessary, avoiding unnecessary checkpoints that would degrade performance while ensuring checkpoints occur when data loss prevention is critical.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The system performs preliminary assessment of checkpoint necessity before initiating checkpointing operations. By evaluating current system state, data change patterns, and network conditions in advance, the system prepares checkpointing only when truly needed, reducing unnecessary processing overhead while maintaining data loss prevention.

Inventive Principle:
Principle #10Preliminary action

3Adaptability or versatility

If dynamic checkpoint interval adjustment is implemented, then adaptability to varying network conditions is improved, but system complexity increases

Engineering Contradiction:
Improveadaptability to network conditionsVSAvoidsystem complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The checkpointing system becomes self-regulating by automatically monitoring its own state and adjusting checkpoint intervals without external intervention. The system services itself by detecting network conditions, evaluating data changes, and autonomously determining optimal checkpoint timing, reducing the need for complex external control mechanisms.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP3090344B1Dynamic checkpointing systems and methods
Publication Date: 2018.07.18 STRATUS TECH BERMUDA LTD
  • EP3090344B1 patent drawingFigure 1
  • EP3090344B1 patent drawingFigure 2(a)~2(b)
  • EP3090344B1 patent drawingFigure 3(a)~3(b)

AI summary

A method for determining a delay in a dynamic, event driven, checkpoint interval. In one embodiment, the method includes the steps of determining the number of network bits to be transferred; determining the target bit transfer rate; calculating the next cycle delay as the number of bits to be transferred divided by the target bit transfer rate. In another aspect, the invention relates to a method for delaying a checkpoint interval. In one embodiment, the method includes the steps of monitoring the transfer of a prior batch of network data and delaying a subsequent checkpoint until the transfer of a prior batch of network data has reached a certain predetermined level of completion. In another embodiment, the predetermined level of completion is 100%.