RAS Service Framework Failover via BMC Core Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Runtime System Management Mode Interrupts (SMIs) cause performance variability and downtime in CPU operations, particularly due to the time-intensive processes of saving and resuming states, and BMC reset latency can lead to prolonged RAS runtime downtime.
Innovation Solution
Implementing a resilient RAS runtime service framework that executes on multiple core partitions with independent exception and reset domains, allowing primary and secondary threads to operate independently, with synchronization through heart-beat signals, to ensure continuous RAS runtime functionality even if one core fails.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If Runtime SMI is used to perform corrective action for RAS events, then RAS service can be provided, but CPU performance variability increases due to service stalls and halts
Solution Approach 1:
The patent extracts RAS runtime services from the host CPU by implementing them in the BMC's out-of-band processing environment. This separation allows RAS services to be provided without causing CPU service stalls or performance variability, as the BMC operates independently from the host CPU execution flow.
2Productivity
If RAS services are offloaded to BMC, then CPU interrupt stalls are reduced, but BMC reset latency causes prolonged RAS runtime downtime
Solution Approach 1:
The patent implements multiple independent RAS runtime service instances across different BMC cores or processor partitions, each with independent exception and reset domains. This local independence ensures that if one instance fails or is reset, other instances continue to provide RAS services without interruption.
Solution Approach 2:
The patent establishes redundant RAS service instances in advance across multiple BMC cores or partitions. When a failure or reset occurs, the system immediately switches to a pre-prepared standby instance, providing failover protection before service interruption can occur.
3Device complexity
If single BMC core executes RAS runtime services, then device complexity is reduced, but single point of failure risk increases
Solution Approach 1:
The patent segments the RAS runtime service execution across multiple BMC cores or processor partitions, with each segment having independent exception and reset domains. This segmentation creates multiple independent failure domains, reducing the impact of any single point of failure while maintaining manageable complexity through modular architecture.
Data Source
AI summary
Examples described herein relate to execution of multiple Reliability Availability Serviceability (RAS) processes on different processors of the at least two processors to provide fallback from a first RAS process to a second RAS process executing on a processor of the at least two processors based on failure or timeout of the first RAS process. In some examples, the different processors comprise independently operating processors whereby failure or inoperability of one of the different processors is independent of another of the different processors. In some examples, failure or timeout of the first RAS process comprises failure of the second RAS process to receive an operating status signal from the first RAS process.