Storage Controller Fencing for Shared Volume Integrity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In shared-resource computing environments, host-initiated operations can lead to data corruption or undetected data loss if a shared volume remains accessible to other host systems during transformations, as these systems may attempt to write to altered data sets, causing directory corruption or losing the effects of the operation.
Innovation Solution
A computer program product and method that involves a storage controller receiving a fencing command to identify and fence other host systems connected to a shared volume, preventing them from accessing it until they refresh their view and validate the changes, thereby maintaining volume integrity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the shared volume remains accessible to other host systems during transforming operations, then availability and productivity are maintained, but data corruption and loss occur
Solution Approach 1:
The storage controller performs preliminary fencing of other host systems before the transforming operation begins. By setting the fenced indicator in advance and notifying host systems prior to the operation, the controller ensures that no other host can access the volume during the transformation, preventing data corruption while allowing the initiating host to proceed with the operation.
2Reliability
If the shared volume is fenced during transforming operations, then data integrity is maintained, but availability and productivity decrease
Solution Approach 1:
The fencing mechanism applies selective access control rather than blanket unavailability. The initiating host system retains full access to the shared volume during transforming operations, while only other host systems are fenced. This localized differentiation allows the operation to proceed with high availability for the initiating host while ensuring data integrity through exclusion of other hosts.
3Productivity
If multiple host systems access the shared volume simultaneously, then availability is maintained, but conflicts and data loss occur
Solution Approach 1:
The storage controller acts as an intermediary that mediates access to the shared volume. It receives the fencing command from the initiating host, determines other connected hosts, and enforces exclusive access by setting fenced indicators and sending notifications. This intermediary control resolves the conflict between concurrent access and data consistency by coordinating access rights through the controller.
4Reliability
If the shared resource is varied offline to one host, then data integrity is ensured, but the operation becomes expensive and time-consuming
Solution Approach 1:
The storage controller performs preliminary fencing of other host systems before the transforming operation begins. By setting the fenced indicator in advance and notifying host systems prior to the operation, the controller ensures that no other host can access the volume during the transformation, preventing data corruption while allowing the initiating host to proceed with the operation.
Data Source
AI summary
A computer-implemented method includes receiving, utilizing a processor, a fencing command at a storage controller from an initiating host system. The fencing command identifies a shared volume. Additionally, the computer-implemented method includes determining which other host systems are connected to the shared volume, and fencing each of the other host systems connected to the shared volume by setting a fencing indicator for each of the other host systems. Still yet, the computer-implemented method includes sending a notification to each of the other host systems that have been fenced, and receiving an unfencing command from at least one of the other host systems that have been fenced.