Split Mode Operation in Fault-Tolerant Computer Systems

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing fault-tolerant computer systems face challenges in rapidly transitioning from redundant operation to split mode without rebooting, maintaining processor state, and allowing continuous service during splitting, while also enabling software upgrades and monitoring.

Innovation Solution

The system design includes designating an active and upgrade subsystem, isolating components, and splitting the lockstep system to maintain identical operational states, allowing independent operation without rebooting, using specific circuitry to preserve the state and regulate lockstep operation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If existing split mode systems are used, then subsystems can operate independently, but the splitting process is difficult and slow requiring reboots

Engineering Contradiction:

Improveindependent subsystem operationVSAvoidsplitting duration

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The system performs preliminary synchronization of subsystems in lockstep mode before splitting, ensuring both subsystems are in identical operational states. This preliminary preparation allows the split to occur rapidly without requiring reboots or lengthy initialization sequences, as the subsystems are already prepared with matching software states.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system dynamically transitions between lockstep mode (for synchronization and fault tolerance) and split mode (for independent operation) through software-controlled switching. This dynamic mode change allows the system to adapt between redundancy and independence based on operational needs without physical reconfiguration or reboots.

Inventive Principle:

Principle #15Dynamics

2Adaptability or versatility

If existing split mode systems are used, then subsystems can operate independently, but processor state is not maintained and reboots are required

Engineering Contradiction:

Improveindependent subsystem operationVSAvoidprocessor state maintenance

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

Before splitting, the system executes a synchronization phase where both subsystems run identical instructions in lockstep mode, ensuring processor states are identical. This preliminary state equalization guarantees that when the split occurs, both subsystems maintain valid, matching processor states without requiring reboots.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system creates a precise copy of the operational state from the active subsystem to the standby subsystem through lockstep execution. This copying mechanism ensures that both subsystems have identical processor states, memory contents, and software configurations before splitting, enabling independent operation without state loss.

Inventive Principle:

Principle #26Copying

3Adaptability or versatility

If existing split mode systems are used, then subsystems can be separated, but service is interrupted for significant periods

Engineering Contradiction:

Improvesubsystem separationVSAvoidcontinuous service capability

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The system performs all necessary synchronization and state preparation while continuing to provide service through the active subsystem. This preliminary action occurs in parallel with service operations, ensuring that when the split occurs, service can continue uninterrupted without significant downtime.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system maintains continuous service capability during the split process by allowing the active subsystem to continue processing while the standby subsystem prepares for independent operation. The useful action of service provision continues without interruption throughout the synchronization and splitting process.

Inventive Principle:

Principle #20Continuity of useful action

4Reliability

If lockstep operation is used, then fault tolerance is achieved through redundancy, but system complexity increases

Engineering Contradiction:

Improvefault toleranceVSAvoidsynchronization mechanism

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The system divides the computer system into separate, identical subsystems that can operate independently or in lockstep. This segmentation allows fault tolerance through redundancy while managing complexity by creating modular, identical units that follow the same execution path, simplifying the synchronization mechanism.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system changes the operational parameter of the subsystems between lockstep mode (synchronized execution) and split mode (independent execution). This parameter change allows the same hardware architecture to provide both fault tolerance through redundancy and simplified operation modes, managing complexity through configurable behavior rather than complex hardware.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS7669073B2Systems and methods for split mode operation of fault-tolerant computer systems

Publication Date: 2010.02.23 STRATUS TECH IRELAND LTD

AI summary

Methods and systems are provided by which a computer system, and in particular, a lockstep fault-tolerant computer system, may be split into a plurality of independently operational subsystems. Each subsystem may be examined, managed or upgraded by an administrator while the overall computer system continues to service end-users. Finally, the separate subsystems may be merged in an efficient fashion and fault-tolerant operation will resume upon the combined system.