your observation is correct. backup node will also download.
If you look at the journey/evolution of hadoop, we had primary, backup
only, checkpointing node and then a generic secondary node.
checking node will do the merge of fsimage and edits
On 25/9/17 5:57 pm, Chang.Wu wrote:
From the official document of Backup Node
<https://hadoop.apache.org/docs/r2.7.2/hadoop-project-dist/hadoop-hdfs/HdfsUserGuide.html#Backup_Node>??
it Says:
The Backup node does not need to download fsimage and edits files from
the active NameNode in order to create a checkpoint, as would be
required with a Checkpoint node or Secondary NameNode, since it
already has an up-to-date state of the namespace state in memory.
I noticed the keyword \'*does not need*\' ??it gives no clear
information about whether or not the Backup Node will download fsimage
and edits files from the active NameNode like checkpoint node.
I checked the source code , namely
|org.apache.hadoop.hdfs.server.namenode.Checkpointer|?? I find that
Both Backup Node and Checkpoint Node will download fsimage and edits
files from active NameNode??no difference.
So , although it is unnecessary for Backup Node to download downlaod
fsimage and edits files from the active NameNode because it keeps a
journal stream from the active name node, but in fact Backup Node
still do so. Right?Thanks.
