Delayed Error Processing in Containerized Systems

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In multi-containerized data centers, a single uncorrectable error in one container can cause failure of the underlying operating system, leading to data loss across other containers, especially when lower priority containers with less robust programming models interact with higher priority, more robust containers, posing challenges in error recovery without disrupting critical services.

Innovation Solution

Implementing delayed error handling, where containers notify other containers to reach a recoverable state before error processing, allowing continued operation and minimizing data loss, with options for shutdown and restart based on system capabilities and user-defined policies, while isolating affected resources to prevent data corruption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If immediate error processing is performed when an uncorrectable error occurs in one container, then the error is handled quickly, but other containers may lose data or be forced to shut down due to operating system failure

Engineering Contradiction:
Improveerror recoveryVSAvoiddata loss
Core Design Contradiction:
ReliabilityVSLoss of information

Solution Approach 1:

The patent applies preliminary action by notifying other containers of the impending error processing in advance, allowing them to complete current tasks and reach a recoverable state before the error is actually processed. This prevents data loss by ensuring containers are in a stable state before error handling occurs, resolving the contradiction between quick error handling and data preservation.

Inventive Principle:
Principle #10Preliminary action

2Stability of the object's composition

If the operating system shuts down immediately upon detecting an uncorrectable error, then system stability is maintained, but critical services in other containers are disrupted

Engineering Contradiction:
Improvesystem stabilityVSAvoidservice availability
Core Design Contradiction:
Stability of the object's compositionVSProductivity

Solution Approach 1:

The patent segments the error handling process by isolating the affected container from the rest of the system. Only the container with the uncorrectable error is processed, while other containers continue to operate normally. This segmentation allows system stability to be maintained for the affected component while preserving service availability for unaffected components.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary notification mechanism that mediates between the error-detecting container and other containers. This intermediary allows the system to prepare other containers for the error processing event, enabling a controlled shutdown or continuation of services rather than an immediate forced termination, thus maintaining both stability and productivity.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Loss of information

If delayed error handling is implemented to allow other containers to reach a recoverable state, then data loss is minimized, but error processing time is increased

Engineering Contradiction:
Improvedata lossVSAvoiderror processing time
Core Design Contradiction:
Loss of informationVSLoss of time

Solution Approach 1:

The patent applies partial action by implementing delayed error handling only for non-critical errors or when containers can reach a recoverable state quickly. For critical errors or when containers cannot recover promptly, immediate error processing is performed. This selective approach minimizes data loss where possible while avoiding excessive delays in critical scenarios.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP3432147B1Delayed error processing
Publication Date: 2021.03.24 INTEL CORP
  • EP3432147B1 patent drawingFigure 1
  • EP3432147B1 patent drawingFigure 2
  • EP3432147B1 patent drawingFigure 3

AI summary

A computing apparatus, including: a hardware platform including a processor and memory; and a system management interrupt (SMI) handler; first logic configured to provide a first container and a second container via the hardware platform; and second logic configured to: detect an uncorrectable error in the first container; responsive to the detecting, generate a degraded system state; provide a degraded state message to the SMI handler; instruct the second container to seek a recoverable state; determine that the second container has entered a recoverable state; and initiate a recovery operation.