I read that in Hadoop 1.0 HDFS has Name Node, Secondary Name Node and Data Nodes. But Secondary Name Node doesn't replace main Name Node if it will fall cause it do some other routine(like stores information to fsimage). In Hadoop 2.0 it is also Name Node, secondary Name Node and Data Nodes. Does the fact that in Hadoop 2.0 secondary Name Node start from lower s means, that this is just substitute Name Node or it is also stores info in fsimage? If yes, why Hadoop 2.0 protected from single point of failure problem?
1 Answers
1
votes
Hadoop 2.0 provides you an option to run two NameNodes in high availability mode, one active and one - hot standby. From HDFS High Availability:
Prior to Hadoop 2.0.0, the NameNode was a single point of failure (SPOF) in an HDFS cluster.
...
The HDFS High Availability feature addresses the above problems by providing the option of running two (or more, as of Hadoop 3.0.0) redundant NameNodes in the same cluster in an Active/Passive configuration with a hot standby(s).
HA requires you to configure additional components for failover and fencing.
Without HA, its the same old NameNode + Secondary NameNode as in HDFS 1.0.
