Tagged articles

JournalNode

2 articles · Page 1 of 1

Sep 8, 2026 · Big Data

HDFS HA Architecture: Components, ZKFC Failover, and Split-Brain Prevention

This article explains HDFS High Availability architecture, detailing core components, the ZKFC-driven automatic failover process, split-brain causes like network partitions and GC pauses, and prevention mechanisms including JournalNode quorum, ZooKeeper locks, and fencing, plus production best practices.

FencingHDFSJournalNode

0 likes · 7 min read

HDFS HA Architecture: Components, ZKFC Failover, and Split-Brain Prevention

Data Thinking Notes

Nov 29, 2022 · Big Data

Understanding HDFS High Availability: Roles, Metadata Persistence, and Failover

This article explains the core concepts of HDFS High Availability, detailing primary and standby NameNode roles, failover mechanisms, shared storage systems, metadata persistence via EditLog and FsImage, and the processes for merging and synchronizing data across active and standby nodes.

EditLogFsImageHDFS

0 likes · 8 min read

Understanding HDFS High Availability: Roles, Metadata Persistence, and Failover