Cluster File System Mounting Verification via Quorum Nodes

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In clustered file systems, there is a challenge in preventing a node from unfencing itself and mounting a file system subsequent to a communication failure, which can lead to data corruption due to potential hardware or software failures.

Innovation Solution

A method is implemented where a cluster manager node is elected to manage disk I/Os, and before a node mounts a file system, it sends a Remote Procedure Call (RPC) to the cluster manager for verification of its active status. The cluster manager queries quorum nodes to confirm its status, ensuring that if communication is lost, the node does not mount the file system, thereby preventing data corruption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a node sends an RPC to the cluster manager for verification before mounting a file system, then data integrity is improved, but the mounting process time increases due to additional verification steps

Engineering Contradiction:
Improvedata integrityVSAvoidmounting process time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The cluster manager pre-establishes a quorum of trusted nodes before any file system mounting operations occur. When a node requests to mount a file system, the verification process quickly queries these pre-selected quorum nodes rather than performing comprehensive system-wide verification, thus maintaining data integrity while reducing mounting time

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements a feedback mechanism where quorum nodes continuously report the cluster manager's status. When verifying a node's request to mount a file system, the cluster manager checks current feedback from quorum nodes to confirm its own active status, ensuring data integrity through real-time status verification without requiring full system re-validation

Inventive Principle:
Principle #23Feedback

2Reliability

If the cluster manager queries quorum nodes to confirm its status before responding, then reliability of file system mounting is improved, but the response time to node requests increases

Engineering Contradiction:
Improvefile system mounting reliabilityVSAvoidresponse time
Core Design Contradiction:
ReliabilityVSSpeed

Solution Approach 1:

Instead of querying all nodes in the cluster to verify status, the cluster manager queries only a partial quorum (a subset) of pre-designated quorum nodes. This partial verification provides sufficient confidence in the cluster manager's status while significantly reducing the time required compared to comprehensive verification of all nodes

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The quorum nodes are pre-selected and configured before any file system operations occur. When the cluster manager needs to verify its status, it queries these pre-established quorum nodes rather than dynamically selecting verification nodes, reducing the overhead and response time of the verification process

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP2176756B1File system mounting in a clustered file system
Publication Date: 2011.11.16 INTERNATIONAL BUSINESS MACHINE CORPORATION
  • EP2176756B1 patent drawingFigure 1
  • EP2176756B1 patent drawingFigure 2

AI summary

A method effectively preventing a requesting node from unfencing and mounting a file system subsequent to a failure in a cluster file system having a plurality of active nodes. The method comprising first upgrading one active node in the cluster to function as a cluster manager node. The cluster manager is in communication with all nodes. The cluster manager is assigned manager responsibilities, in part, comprising first receiving an active status request from the node requesting to mount a file system. The cluster manager first queries the quorum nodes to determine whether each node considers the cluster manager to still have cluster management responsibilities for the file system. If a majority of quorum nodes consider the cluster manager to still have cluster management responsibilities for the file system then the cluster manager responds to the requesting node's active status request. Thereafter, the requesting node proceeds with mounting the file system.