|
gptkbp:instanceOf
|
data storage technique
|
|
gptkbp:alternativeTo
|
3x replication
|
|
gptkbp:appliesTo
|
files
directories
|
|
gptkbp:canBe
|
HDFS storage policies
|
|
gptkbp:canBeDisabled
|
administrator
|
|
gptkbp:combines
|
HDFS snapshots
|
|
gptkbp:compatibleWith
|
HDFS encryption zones (in some versions)
|
|
gptkbp:defaultErasureCodingPolicy
|
RS-6-3-1024k
|
|
gptkbp:enables
|
efficient storage utilization
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
Erasure Coding in HDFS
|
|
gptkbp:improves
|
latency for small reads/writes
|
|
gptkbp:introducedIn
|
gptkb:Hadoop_3.0.0
|
|
gptkbp:notEnabledByDefault
|
in Hadoop 3.x
|
|
gptkbp:notRecommendedFor
|
hot data
|
|
gptkbp:parityAlgorithm
|
gptkb:Reed-Solomon
|
|
gptkbp:policy
|
RS-6-3-1024k
|
|
gptkbp:purpose
|
reduce storage overhead
|
|
gptkbp:recoveryMethod
|
data node failures
|
|
gptkbp:recoveryProcess
|
decoding from data and parity blocks
|
|
gptkbp:reduces
|
storage cost
|
|
gptkbp:requires
|
more CPU resources
HDFS version 3.0.0 or later
minimum number of data nodes
|
|
gptkbp:savesStorage
|
up to 50%
|
|
gptkbp:splitsDataInto
|
data blocks
parity blocks
|
|
gptkbp:supportedBy
|
gptkb:HDFS_API
HDFS CLI
|
|
gptkbp:tradeoff
|
higher CPU and network usage
|
|
gptkbp:usedBy
|
large-scale data storage systems
|
|
gptkbp:usedFor
|
cold data
|
|
gptkbp:usedIn
|
gptkb:Hadoop_Distributed_File_System
|
|
gptkbp:bfsParent
|
gptkb:Hadoop_3.x
|
|
gptkbp:bfsLayer
|
8
|