Host Anomaly Snapshots for Unscheduled Shutdown Data Integrity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In data centers, unscheduled shutdowns of host servers often result in unsynchronized or corrupted data due to the lack of data flushing to storage nodes, and storage objects may be targeted for attack during reboot, especially in cases of malware-induced shutdowns.
Innovation Solution
Implementing a system that detects the type of shutdown (scheduled or unscheduled) and generates an emergency snapshot of storage objects only when an unscheduled shutdown occurs, using host-resident software and a storage array to manage snapshot generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If data is not flushed to storage node during unscheduled shutdown, then shutdown speed is improved, but data consistency deteriorates
Solution Approach 1:
The system performs preliminary actions by detecting the shutdown type (scheduled vs. unscheduled) and conditionally generating snapshots before the reboot process begins. This preliminary detection and conditional snapshot creation ensures data protection is in place before any potential corruption occurs, resolving the contradiction by preparing protective measures in advance only when necessary.
Solution Approach 2:
The system changes the operational parameter of snapshot generation based on the shutdown type detection. For scheduled shutdowns, snapshot generation is skipped (maintaining speed), while for unscheduled shutdowns, snapshot generation is activated (ensuring reliability). This parameter-based conditional execution resolves the contradiction by adapting the system behavior to the specific shutdown scenario.
2Reliability
If emergency snapshot is generated for every unscheduled shutdown, then data integrity is improved, but system complexity increases
Solution Approach 1:
The system applies local quality by generating snapshots selectively only for storage objects associated with hosts that experienced unscheduled shutdowns, rather than creating snapshots for all storage objects in the system. This localized approach maintains data integrity for affected areas while avoiding unnecessary complexity and resource consumption in unaffected areas.
Solution Approach 2:
The system implements self-service by automatically detecting unscheduled shutdowns and triggering snapshot generation without requiring manual intervention or complex external coordination. The storage system monitors host status, identifies anomalies, and autonomously creates protective snapshots, reducing overall system complexity while maintaining reliability.
3Measurement precision
If shutdown type detection is implemented, then data protection accuracy is improved, but processing time increases
Solution Approach 1:
The system performs shutdown type detection as a preliminary action during the reboot process, leveraging information that becomes available when the host system restarts. By detecting the shutdown type early in the reboot sequence before normal operations resume, the system achieves accurate classification without significantly impacting overall processing time.
Solution Approach 2:
The system skips detailed analysis for scheduled shutdowns (which are easily identifiable and low-risk) and rushes through the detection process, only performing comprehensive snapshot generation for clearly identified unscheduled shutdowns. This selective processing approach maintains detection accuracy while minimizing time loss for routine scenarios.
Data Source
AI summary
Host-resident software such as a Multi-Path Input-Output (MPIO) driver generates notifications of scheduled host shutdowns and unscheduled host shutdowns. A storage node generates emergency snapshots of storage objects masked to the host server in response to receipt of an unscheduled shutdown notification. The storage node also generates emergency snapshots of storage objects masked to a host server in response to detection of a reboot sequence of the host server when a corresponding scheduled shutdown notification was not received beforehand.