GFS

E702632

GFS is a scalable, fault-tolerant distributed file system developed by Google to efficiently manage large data-intensive applications across many commodity servers.

Try in SPARQL Jump to: Statements Referenced by

Statements (50)

Predicate Object
instanceOf Google technology
distributed file system
fault-tolerant file system
scalable file system
architecture master–chunkserver architecture
consistencyModel relaxed consistency with record append semantics
dataUnit chunk
designedFor batch processing workloads
large data-intensive applications
large-scale data processing
developer Google
fullName Google File System NERFINISHED
handlesFailureMode chunkserver failures
disk failures
network partitions
influenced Colossus NERFINISHED
Hadoop Distributed File System NERFINISHED
lessOptimisedFor small random writes
license proprietary
metadataIncludes chunk locations
file namespace
file-to-chunk mapping
metadataStoredBy master server
optimisedFor large sequential reads
large sequential writes
paperAuthors Howard Gobioff NERFINISHED
Sanjay Ghemawat NERFINISHED
Shun-Tak Leung NERFINISHED
paperPublishedAt SOSP 2003 NERFINISHED
paperTitle The Google File System NERFINISHED
replacedBy Colossus NERFINISHED
replicationFactorDefault 3
runsOn commodity servers
scopeOfDeployment internal to Google
supportsFeature atomic record append
automatic re-replication
data replication
fault tolerance
high throughput
rebalancing of chunks
recovery from chunkserver failures
relaxed consistency model
snapshot
typicalChunkSize 64 MB
usedBy Google MapReduce NERFINISHED
Google data analysis pipelines
Google search infrastructure NERFINISHED
usesComponent chunkserver
client library
master server

Referenced by (1)

Full triples — surface form annotated when it differs from this entity's canonical label.