Power Operation Controller Process Migration for Cluster Availability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In information handling systems (IHS) with clustered environments, power operations such as reboots can inadvertently interrupt delegated processes, leading to loss of availability and file corruption due to insufficient redundancy and lack of multipath data transfer, compromising high availability requirements.
Innovation Solution
A power operation controller is implemented to determine power operation requests and transfer delegated processes to another network resource, preventing the IHS from performing power operations during the transfer, thus ensuring continuous operation and data integrity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a server in a cluster performs a power operation (reboot/shutdown), then the server can be maintained or updated, but the delegated processes running on that server are interrupted causing loss of availability and potential file corruption
Solution Approach 1:
The system performs preliminary detection of power operation requests and initiates process migration before the actual power operation occurs. The method detects a power operation request, determines that a delegated process is running, migrates the process to another node, and only then allows the power operation to proceed, preventing service interruption.
Solution Approach 2:
The system introduces an intermediary control mechanism that mediates between the power operation request and the running delegated process. This intermediary detects the conflict and coordinates the migration of the process to another node before the power operation executes, acting as a buffer to prevent direct interruption.
2Speed
If the system allows power operations to proceed immediately upon request, then power management responsiveness is improved, but process migration cannot be completed leading to service unavailability
Solution Approach 1:
The system performs the process migration action preliminarily before executing the power operation. By detecting the power operation request first and completing the migration in advance, the system ensures that when the power operation executes, the delegated process has already been relocated, maintaining service availability while preserving power management responsiveness.
Solution Approach 2:
The system dynamically adjusts the execution timing of power operations based on the state of delegated processes. Rather than executing power operations immediately or on a fixed schedule, the system monitors for running delegated processes and delays power operation execution until migration is complete, making the power management system adaptive to current operational conditions.
3Reliability
If the system implements process migration before power operations, then service continuity is maintained, but the complexity of power management increases
Solution Approach 1:
The system implements self-service by having the delegated process automatically detected and migrated when a power operation is requested. The process manager autonomously identifies the conflict between the power operation and running process, initiates migration to another node, and only then permits the power operation, reducing the need for complex external coordination.
Solution Approach 2:
The system uses feedback mechanisms where the state of delegated processes is continuously monitored and fed back to the power management system. When a power operation request is detected, the system receives feedback about running processes, adjusts its behavior accordingly by delaying the power operation until migration completes, and then confirms service continuity.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An information handling system (IHS), power operation controller, and method provide for execution of delegated processes in a cluster without inadvertent interruption due to a power operation of the IHS. The power operation controller of the IHS has a processing subsystem that is communicatively coupled to a network interface and a memory containing a power operation utility. The processing subsystem executes, via the network interface, a delegated process on a network resource. The processing subsystem executes the power operation utility to cause the IHS to determine whether a request for a power operation of the IHS is received. In response to determining that a power operation is received, the processing subsystem transfers the delegated process to another network resource. In addition, during a transference of the delegated process, the processing subsystem prevents the IHS from performing a power operation of the IHS that would cause interruption of the delegated process.