The Google File System (GFS) is a distributed large-scale file system designed by Google. It is built to handle component failures, store large files, and optimize performance for file appending. The system features a single master, multiple chunkservers, and multiple clients. GFS uses chunk replication, lease management, and checksumming to ensure high availability and data integrity.
Table of contents
AssumptionsInterfaceArchitectureSingle MasterChunk SizeMetadataIn-Memory Data StructuresChunk LocationsOperation LogConsistency ModelGuarantees by GFSImplications for Applications34 Impressions