Snapshot Message Fabric Blockage Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Complex fabric implementations in computer systems and networks often encounter issues like deadlocks and starvation during post-silicon testing, making debugging challenging due to limited visibility and interference from hardware/software error handling.

Innovation Solution

A device with a detection unit and a snapshot unit is implemented to detect blockages by sending a snapshot message along a second channel, allowing nodes to capture the fabric state before corrective actions occur, thereby gathering accurate information about the blockage event.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If complex fabric implementations are used to facilitate fast and reliable communication, then communication speed and reliability are improved, but the system becomes more prone to blockage issues like deadlocks and starvation

Engineering Contradiction:

Improvecommunication reliabilityVSAvoidfabric complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent implements a detection unit that proactively monitors fabric links for blockage conditions before they fully develop into deadlocks or starvation. By detecting potential blockages early through monitoring transmission delays and link states, the system can trigger snapshot capture and debugging mechanisms preemptively, preventing the complex blockage issues from fully manifesting while maintaining the complex fabric architecture's communication reliability

Inventive Principle:

Principle #10Preliminary action

2Reliability

If post-silicon testing is performed on complex fabrics, then real-world performance is validated, but debugging becomes extremely challenging due to limited visibility

Engineering Contradiction:

Improvefabric performance validationVSAvoidblockage detection difficulty

Core Design Contradiction:

ReliabilityVSDifficulty of detecting and measuring

Solution Approach 1:

The patent captures fabric state snapshots at multiple nodes preemptively when blockage conditions are detected, before the blockage fully develops and before error handling mechanisms interfere. This preliminary capture of state information provides visibility into the fabric's actual operating conditions during post-silicon testing, enabling effective debugging while validating real-world performance

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent introduces snapshot messages as intermediary carriers that transport state information from distributed fabric nodes to debugging systems. These snapshot messages serve as intermediaries that bridge the visibility gap in complex fabrics, allowing external debugging tools to observe internal fabric states without interfering with normal operation or requiring direct access to proprietary internal structures

Inventive Principle:

Principle #24Intermediary (Mediator)

3Reliability

If hardware and software error handling is enabled, then system stability is improved, but it interferes with debugging by unintentionally freeing the fabric from blocked states

Engineering Contradiction:

Improvesystem stabilityVSAvoidblockage state information

Core Design Contradiction:

ReliabilityVSLoss of information

Solution Approach 1:

The patent triggers snapshot capture immediately when blockage detection occurs, before error handling mechanisms have a chance to intervene and resolve the blockage. This preliminary action preserves the authentic blockage state information including the distributed node states and link conditions, while still allowing error handling to operate for system stability. The timing ensures that debugging information is captured before it is lost

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent creates copies of the fabric state through snapshot messages that are sent to debugging systems. These copies preserve the blockage state information independently of the original fabric state, which may subsequently be modified by error handling mechanisms. The snapshot copies maintain the historical blockage information for debugging purposes while allowing the system's error handling to restore stability

Inventive Principle:

Principle #26Copying

4Loss of information

If distributed fabric states are captured across multiple nodes, then debugging visibility is improved, but the amount of data to be processed increases

Engineering Contradiction:

Improvefabric state visibilityVSAvoidsnapshot data volume

Core Design Contradiction:

Loss of informationVSQuantity of substance

Solution Approach 1:

The patent divides the fabric state capture into segmented snapshots at individual nodes rather than attempting to capture the entire fabric state simultaneously. Each node captures its local state independently through snapshot messages, creating manageable segments of state information. This segmentation reduces the complexity of data collection while maintaining comprehensive visibility through the distributed nature of the segmented captures

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS9929899B2Snapshot message

Publication Date: 2018.03.27 HEWLETT PACKARD ENTERPRISE DEV LP

AI summary

A blockage is detected at a first link based on a delay and/or stoppage of transmission of a data message along the first link between first and second nodes of a plurality of nodes of a fabric. A snapshot message is sent along at least a second link between the first and second nodes in response to the blockage being detected. The second node may capture a fabric state at the second node in response to receiving the snapshot message, before a corrective action occurs.