ZKFailoverController

E829912

ZKFailoverController is an HDFS high-availability component that uses ZooKeeper to coordinate automatic failover between NameNodes.

Jump to: Statements Referenced by

Statements (40)

Predicate Object
instanceOf HDFS component
ZooKeeper-based failover controller
high-availability controller
belongsToModule Hadoop HDFS NERFINISHED
belongsToProject Apache Software Foundation Hadoop project NERFINISHED
canBeUsedWith manual failover as backup procedure
coordinates automatic failover between NameNodes
dependsOn reliable ZooKeeper quorum
elects active NameNode via ZooKeeper
ensures only one active NameNode at a time
implements automatic failover policy
interactsWith FailoverController framework in Hadoop
NameNode RPC server NERFINISHED
ZooKeeper client library NERFINISHED
isConfiguredIn core-site.xml
hdfs-site.xml
isConfiguredWith fencing methods
health check parameters
isDocumentedIn Hadoop HDFS High Availability documentation
isImplementedIn Java NERFINISHED
isStartedBy hdfs zkfc command
logsTo Hadoop log files
manages active NameNode role
standby NameNode role
monitors NameNode health
partOf Apache Hadoop NERFINISHED
HDFS high availability architecture
performs fencing of previously active NameNode on failover
health monitoring via RPC to NameNode
prevents split-brain scenarios between NameNodes
requires ZooKeeper ensemble
properly configured ZooKeeper ACLs
shared edits or Quorum Journal Manager setup
runsAlongside NameNode
runsOn NameNode host
storesCoordinationStateIn ZooKeeper znodes NERFINISHED
supports HDFS NameNode high availability
supportsMode automatic failover
uses Apache ZooKeeper NERFINISHED
wasIntroducedIn Hadoop 2.x high-availability feature set

Referenced by (1)

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

HDFS hasComponent ZKFailoverController