Storage Server Metadata Caching for Takeover Performance

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In clustered storage server systems, the time required for takeover and giveback operations increases with the number of volumes, due to the need for extensive disk I/O operations during these transitions.

Innovation Solution

Implementing a method where each storage server periodically pushes metadata associated with its volumes to its partner server, storing it in the partner's metadata cache, allowing for reduced disk access during takeover and giveback operations by utilizing cached metadata.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If metadata is retrieved from disk subsystem during mount process, then volume mounting can be completed, but the number of disk I/O operations increases significantly

Engineering Contradiction:

Improvevolume mounting completenessVSAvoidnumber of disk I/O operations

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

The system performs preliminary actions by caching metadata from the disk subsystem into memory before actual mount operations are needed. During normal operation, metadata is proactively read and stored in a cache structure, so that when takeover or giveback operations occur, the metadata is already available in memory rather than requiring disk I/O at that critical moment.

Inventive Principle:

Principle #10Preliminary action

2Reliability

If all volume metadata is accessed from disk during takeover and giveback operations, then complete volume mounting is achieved, but the time required to complete these operations increases

Engineering Contradiction:

Improvevolume mounting completenessVSAvoidtime required for takeover and giveback operations

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

Metadata is proactively cached into memory during normal operation before takeover or giveback scenarios occur. This preliminary caching ensures that when these time-critical operations are needed, the metadata is already available in fast memory rather than requiring slow disk access, thereby reducing operational time while maintaining completeness.

Inventive Principle:

Principle #10Preliminary action

3Adaptability or versatility

If the number of volumes hosted by storage servers increases, then storage capacity and functionality are improved, but the number of disk I/O operations required for mounting increases

Engineering Contradiction:

Improvestorage capacityVSAvoidnumber of disk I/O operations

Core Design Contradiction:

Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

As volumes are added to the storage system, their metadata is automatically cached into memory during normal operation. This preliminary caching scales with the number of volumes, ensuring that even as storage capacity increases, the metadata for all volumes is pre-loaded into fast memory, preventing disk I/O operations from increasing linearly with volume count during critical operations.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS7836020B1Method and apparatus to improve server performance associated with takeover and giveback procedures

Publication Date: 2010.11.16 NETAPP INC

AI summary

A system is provided to improve storage server performance. The system comprises a receiving component of a storage server to receive a partner server's file system metadata while the partner server operates in a normal mode. The partner server's file system metadata is stored in the storage servers metadata cache and is used to mount the partner server's file system when the first server transitions into a takeover mode of operation.