Lightweight Filesystem Cache for Remote Immutable Data Access

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Modern general-purpose filesystems incur overhead for database systems that operate on large, immutable files stored remotely, as they maintain unnecessary data structures like inodes and journals, affecting performance.

Innovation Solution

A lightweight filesystem that divides local storage into data blocks with metadata in the last portion, eliminating directory hierarchies and using remote data store identifiers, allowing metadata to be memory-efficient and cacheable.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If a general-purpose filesystem is used to manage remote storage, then data access functionality is provided, but computing resource usage increases due to overhead from maintaining inodes, journals, and directory hierarchies

Engineering Contradiction:
Improvedata access performanceVSAvoidcomputing resource usage
Core Design Contradiction:
ProductivityVSUse of energy by moving object

Solution Approach 1:

The patent extracts and removes unnecessary filesystem components (inodes, journals, complex directory hierarchies) from the system, keeping only the essential metadata needed for remote storage access. This eliminates the overhead of maintaining general-purpose filesystem structures while preserving core data access functionality.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

Instead of using a full filesystem structure to manage remote storage, the patent inverts the approach by using a simplified metadata structure that directly maps to remote storage objects, eliminating the need for traditional filesystem abstractions.

Inventive Principle:
Principle #13The other way round (Inversion)

2Adaptability or versatility

If a general-purpose filesystem with directory hierarchies is used, then file organization capability is provided, but device complexity increases due to unnecessary data structures

Engineering Contradiction:
Improvefile organization capabilityVSAvoidfilesystem structure complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent removes complex directory hierarchy structures and replaces them with a flat metadata structure that directly references remote storage objects, eliminating unnecessary organizational layers while maintaining file access capability.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent segments the filesystem into minimal essential components only - primarily metadata structures that reference remote storage - eliminating the need for complex hierarchical organization while preserving basic file access functionality.

Inventive Principle:
Principle #1Segmentation

3Reliability

If sophisticated data structures like inodes and journals are maintained, then crash recovery and access control features are provided, but access performance deteriorates due to overhead

Engineering Contradiction:
Improvecrash recovery capabilityVSAvoiddata access speed
Core Design Contradiction:
ReliabilityVSSpeed

Solution Approach 1:

The patent extracts and removes journaling structures and complex inode maintenance mechanisms, relying instead on the simplicity of the metadata structure and remote storage properties to provide adequate recovery capability without the performance overhead of traditional filesystem safeguards.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentUS20260050579A1Lightweight filesystem for remote storage caching
Publication Date: 2026.02.19 AMAZON TECH INC
  • US20260050579A1 patent drawing
  • US20260050579A1 patent drawing
  • US20260050579A1 patent drawing

AI summary

A lightweight filesystem may be provided for remote storage caching. A filesystem may maintain a persistent cache for a data set stored as data files in immutable data objects in a remote data store. Filesystem metadata may be evaluated to determine whether a portion of a data file is stored in the persistent cache according to an offset and length specified in a request. If in the persistent cache, data obtained from a data block in the persistent cache may be returned. If not in the persistent cache, then the remote data store may be accessed and the data file in the immutable data object read to obtain the portion of the data file.