I used following property for reduce dead node timeout.
Propertyname : dfs.heartbeat.recheck.interval
value : 1
But when I remove datanode from cluster this details not removed from hadoop cluster.It is in dead node state only in that cluster.
Please suggest any way to remove removed datanode details from hadoop cluster.
You can view the live nodes or dead nodes alone by using the below HDFS commands
hdfs dfsadmin -report -live
hdfs dfsadmin -report -dead
You can get the live node name or any other particular details using the below HDFS command
hdfs dfsadmin -report -live | grep Name:
Hope it helps.
Related
I have a cluster of 5 machines:
1 big NameNode
4 standard DataNodes
I want to change my current NameNode with a DataNode without losing the data stored in HDFS, so my cluster could become:
1 standard NameNode
3 standard DataNodes
1 big DataNode
Does someone know a simple way to do that?
Thank you very much
Decomission data node where namenode will be moved.
Stop the cluster.
Create a tar of dfs.name.dir from current namenode.
Copy all hadoop config files from current NN to target NN.
Replace the name/ip of target namenode by modifying core-site.xml.
Restore tarball of dfs.name.dir. Make sure that full path is same.
Now start the cluster by starting new namenode and one less datanode.
Verify that everything is working perfectly.
Add old namenode as datanode by configuring it as datanode.
I would suggest to uninstall and then install hadoop on both the nodes so that previous configuration does not cause any problem.
I enabled the Namenode High Availability using ambari.
I want to verify the connection using dfs.nameservices (nameservice ID) before start the coding.
Is there any command line or tool to verifiy it?
You can use the normal HDFS CLI.
hdfs dfs -ls hdfs://nameservice/user
Which should also work the same as
hdfs dfs -ls hdfs:///user
Or giving your active namenode
hdfs dfs -ls hdfs://namenode-1:port/user
If you provide the standby namenode, it will say operation READ not supported in state standby
In Hadoop HDFS Federation the latest step of adding a new NameNode to an existing HDFS cluster is:
==> Refresh the Datanodes to pickup the newly added Namenode by running the following command against all the Datanodes in the cluster:
[hdfs]$ $HADOOP_PREFIX/bin/hdfs dfsadmin -refreshNameNodes <datanode_host_name>:<datanode_rpc_port>
Witch is the best place to execute the flowing command: NameNode or datanode ?
If I have 1000 Datanodes is it logical to run it 1OOO time ?
In namenode run this command once.
$HADOOP_PREFIX/sbin/slaves.sh hdfs dfsadmin -refreshNameNodes <datanode_host_name>:<datanode_rpc_port>
slaves.sh script will distribute the command to all the slave hosts which are mentioned in slaves file (typically placed in $HADOOP_CONF_DIR)
I installed my test cluster using Cloudera Manager free.
I can only browse the filesystem from the main NameNode. When running hadoop dfs -ls only shows the local folder.
JPS shows the Jps, TaskTracker, DataNode on the nodes.
MapReduce tasks/jobs run fine on all the nodes as a cluster.
With my custom setup Hadoop cluster (without Cloudera), I can easily browse and manipulate the hdfs filesystem (eg. I can run hadoop dfs -mkdir test1 on all the nodes - but only on the NameNode in CDH4)
why is this?
Try using the command ./bin/hadoop fs -ls / for HDFS browsing.
I am fairly new to hadoop. For running some benchmarks, I need variety of hadoop configuration for comparison.
I want to know a method to remove a hadoop slave from DFS (not running datanode daemon anymore) but not from Mapred (keep running tasktracker), or vice-versa.
AFAIK, there is a single slave file for such hadoop nodes and not separate slave files for DFS and Mapred.
Currently, I am trying to start both DFS and Mapred on the slave node , and then killing datanode on the slave. But it takes a while to put that node in to 'dead nodes' on HDFS GUI. Any parameter can be tuned to make this timeout quicker ?
Thankssss
Try using dfs.hosts and dfs.hosts.exclude in the hdfs-site.xml, mapred.hosts and mapred.hosts.exclude in mapred-site.xml. These are for allowing/excluding hosts to connect to the NameNode and the JobTracker.
Once the list of nodes in the files has been updated appropriately, the NameNode and the JobTracker have to be refreshed using the hadoop dfsadmin -refreshNodes and hadoop mradmin -refreshNodes command respectively.
Instead of using slaves file to start all processes on your cluster, you can start only required daemons on each machine if you have few nodes.