How to set up a federated cluster? - hadoop

I will paste all my configuration below. I have a cluster of 3 computers. Configuration of namenode 1 (impc2361)
core-site.xml:
<configuration>
<property>
<name>fs.defaultFS</name>
<value>viewfs//ClusterA</value>
</property>
<property>
<name>fs.viewfs.mounttable.ClusterA.link./home</name>
<value>hdfs//impc2361:8021/home</value>
</property>
<property>
<name>fs.viewfs.mounttable.ClusterA.link./home1</name>
<value>hdfs//impc2359:8020/home1</value>
</property>
<property>
<name>hadoop.tmp.dir</name>
<value>/home/hadoop/Downloads/hadoop2/tmpfold</value>
</property>
</configuration>
hdfs-site.xml:
<configuration>
<property>
<name>dfs.replication</name>
<value>1</value>
</property>
<property>
<name>dfs.nameservices</name>
<value>home,home1</value>
</property>
<property>
<name>dfs.namenode.rpc-address.home</name>
<value>impc2361:8021</value>
</property>
<property>
<name>dfs.namenode.http-address.home</name>
<value>impc2361:50070</value>
</property>
<property>
<name>dfs.namenode.rpc-address.home1</name>
<value>impc2359:8020</value>
</property>
<property>
<name>dfs.namenode.http-address.home1</name>
<value>impc2359:50070</value>
</property>
</configuration>
I have copied the same configurations on the other nodes as well that is namenode2 (impc2359) and datanode (impc2391)
Problems
I don't get the web^page of namenode1(impc2361) when I type impc2361.htcitmr:50070 in web url
It throws an error
HTTP ERROR 404
Problem accessing /dfshealth.jsp.
Reason: NOT_FOUND
I get a web page of namenode2 (impc2359) when I type impc2359.htcitmr:50070 but i don't find the folder /home1 which was set in core-site.xml
I am not able to do any operations through my terminal on cluster as it throws a error that it is readonly
hadoop fs -mkdir /a
mkdir: InternalDir of ViewFileSystem is readonly; operation=mkdirsPath=/a
Please kindly help

Related

No Disk Space allocated to HDFS filesystem

I am trying to set up Hadoop on my local machine. However, when I'm running wordcount based on map reduce example (I did hdfs namenode -format earlier) :
This is maybe hard to read but I end up with a "Job failed with state FAILED due to:
Application failed 2 times
due to AM Container exited with
exitCode: -1000 Failing this attempt.Diagnostics: No space available in any of the local directories."
I don't understand why I have such error. This is how my applications & attempt look like :
I followed several tutorials, ending up with these parameters :
mapred-site.xml :
configuration>
<property>
<name>mapreduce.framework.name</name>
<value>yarn</value>
</property>
</configuration>
yarn-site.xml :
<configuration>
<!-- Site specific YARN configuration properties -->
<property>
<name>yarn.nodemanager.aux-services</name>
<value>mapreduce_shuffle</value>
</property>
<property>
<name>yarn.nodemanager.auxservices.mapreduce.shuffle.class</name>
<value>org.apache.hadoop.mapred.ShuffleHandler</value>
</property>
<property>
<name>yarn.nodemanager.disk-health-checker.enable</name>
<value>false</value>
</property>
<property>
<name>yarn.application.classpath</name>
<value>
%HADOOP_HOME%\etc\hadoop,
%HADOOP_HOME%\share\hadoop\common\*,
%HADOOP_HOME%\share\hadoop\common\lib\*,
%HADOOP_HOME%\share\hadoop\hdfs\*,
%HADOOP_HOME%\share\hadoop\hdfs\lib\*,
%HADOOP_HOME%\share\hadoop\mapreduce\*,
%HADOOP_HOME%\share\hadoop\mapreduce\lib\*,
%HADOOP_HOME%\share\hadoop\yarn\*,
%HADOOP_HOME%\share\hadoop\yarn\lib\*
</value>
</property>
</configuration>
core-site.xml :
<configuration>
<property>
<name>fs.default.name</name>
<value>hdfs://localhost:9000</value>
</property>
</configuration>
hdfs-site.xml :
<configuration>
<property>
<name>dfs.replication</name>
<value>1</value>
</property>
<property>
<name>dfs.namenode.name.dir</name>
<value>file:///C:/hadoop-3.3.0/data/namenode</value>
</property>
<property>
<name>dfs.datanode.data.dir</name>
<value>file:///C:/hadoop-3.3.0/data/datanode</value>
</property>
</configuration>
Can you help me with this please ? I already tried what is mentionned in these questions :
Hadoop errorcode -1000, No space available in any of the local directories, except for the namenode emptying cache part.
Hadoop Windows setup. Error while running WordCountJob: "No space available in any of the local directories"
what do you think ?
Thank you !

hadoop's start-dfs not creating datanode on the slave

I am trying to set a Hadoop cluster over two nodes. start-dfs.sh on my master node is opening a window and shortly after the window closes, and when i execute start-dfs it logs namenode is correctly launched, but datanode is not and logs the following :
Problem binding to [slave-VM1:9005] java.net.BindException: Cannot assign requested address: bind; For more details see: http://wiki.apache.org/hadoop/BindException
I have set
ssh-keygen -t rsa -P ''
cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys
(and also set adminstrators_authorized_keys file with the right public key) (also ssh user#remotemachine is working and gives access to the slave)
Here's my full Hadoop configuration set on both master and slave machines (Windows):
hdfs-site.xml :
<configuration>
<property>
<name>dfs.name.dir</name>
<value>/C:/Hadoop/hadoop-3.2.2/data/namenode</value>
</property>
<property>
<name>dfs.datanode.https.address</name>
<value>slaveVM1:50475</value>
</property>
<property>
<name>dfs.data.dir</name>
<value>/C:/Hadoop/hadoop-3.2.2/data/datanode</value>
</property>
<property>
<name>dfs.replication</name>
<value>2</value>
</property>
</configuration>
core-site.xml :
<configuration>
<property>
<name>dfs.datanode.http.address</name>
<value>slaveVM1:9005</value>
</property>
<property>
<name>fs.default.name</name>
<value>hdfs://masterVM2:9000</value>
</property>
<property>
<name>hadoop.tmp.dir</name>
<value>/C:/Hadoop/hadoop-3.2.2/hadoopTmp</value>
</property>
<property>
<name>fs.defaultFS</name>
<value>hdfs://masterVM2:8020</value>
</property>
</configuration>
mapred-site.xml
<configuration>
<property>
<name>mapred.job.tracker</name>
<value>masterVM2:9001</value>
</property>
<property>
<name>mapreduce.framework.name</name>
<value>yarn</value>
</property>
<property>
<name>mapreduce.application.classpath</name>
<value>%HADOOP_HOME%/share/hadoop/mapreduce/*,%HADOOP_HOME%/share/hadoop/mapreduce/lib/*,%HADOOP_HOME%/share/hadoop/common/*,%HADOOP_HOME%/share/hadoop/common/lib/*,%HADOOP_HOME%/share/hadoop/yarn/*,%HADOOP_HOME%/share/hadoop/yarn/lib/*,%HADOOP_HOME%/share/hadoop/hdfs/*,%HADOOP_HOME%/share/hadoop/hdfs/lib/*</value>
</property>
</configuration>
yarn-site.xml
<configuration>
<property>
<name>yarn.acl.enable</name>
<value>0</value>
</property>
<property>
<name>yarn.resourcemanager.hostname</name>
<value>master</value>
</property>
<property>
<name>yarn.nodemanager.aux-services</name>
<value>mapreduce_shuffle</value>
</property>
</configuration>
PS : i am adminstrator on both machines, and i set HADOOP_CONF_DIR C:\Hadoop\hadoop-3.2.2\etc\hadoop
I also set the slave IP in hadoop_conf_dir slaves file.
PS : if i remove the code :
<property>
<name>dfs.datanode.https.address</name>
<value>slave:50475</value>
</property>
from hdfs-site.xml
Then both datanote and namenode launch on the master node.
hosts :
*.*.*.* slaveVM1
*.*.*.* masterVM2
... are the IPs of the respective machines, all other entries are commented out
This usually happens
BindException: Cannot assign requested address: bind;
when the port in use. Meaning maybe it's the application was already started, or was started previously and didn't shut down properly or another applicaiton is using that port. Try rebooting, (as a heavy handed but reasonably effective way of clearing ports).

Hadoop jobtracker's tracking url cannot access

I have configured my hadoop system in wsl and run the wordcount example. But when I want to see the history of the job, I found the tracking url cannot access.
The job is working well, the jobhistory is running as well.
The history tracking url is my wsl hostname:8088/proxy/application_1585482453915_0002/.
You can see the url above.
But I can still access to localhost:19888/jobhistory to see my jobhistory.
How is this problem occurs? Is it a problem of configuration?
My hadoop version is 2.7.1.
My core-site.xml
<property>
<name>hadoop.tmp.dir</name>
<value>file:/home/hadoop/hadoop/tmp</value>
<description>Abase for other temporary directories.</description>
</property>
<property>
     <name>fs.defaultFS</name>
     <value>hdfs://localhost:9000</value>
</property>
My hdfs-site.xml
<property>
<name>dfs.replication</name>
<value>1</value>
</property>
<property>
<name>dfs.namenode.name.dir</name>
<value>file:/home/hadoop/hadoop/tmp/dfs/name</value>
</property>
<property>
<name>dfs.datanode.data.dir</name>
<value>file:/home/hadoop/hadoop/tmp/dfs/data</value>
</property>
My mapred-site.xml
<property>
<name>mapreduce.framework.name</name>
<value>yarn</value>
</property>
<property>
<name>mapreduce.jobhistory.address</name>
<value>localhost:10020</value>
</property>
<property>
<name>mapreduce.jobhistory.webapp.address</name>
<value>localhost:19888</value>
</property>
My yarn-site.xml
<property>
<name>yarn.nodemanager.vmem-check-enabled</name>
<value>false</value>
<description>Whether virtual memory limits will be enforced for containers</description>
</property>
<property>
<name>yarn.nodemanager.vmem-pmem-ratio</name>
<value>4</value>
<description>Ratio between virtual memory to physical memory when setting memory limits for containers</description>
</property>
<property>
<name>yarn.nodemanager.aux-services</name>
<value>mapreduce_shuffle</value>
</property>
<property>
<name>yarn.nodemanager.aux-services.mapreduce_shuffle.class</name>
<value>org.apache.hadoop.mapred.ShuffleHandler</value>
</property>
My /etc/hosts
127.0.0.1 localhost
127.0.1.1 DESKTOP-U1EOV4J.localdomain DESKTOP-U1EOV4J
The JobHistoryServer daemon is running in localhost (127.0.0.1), whereas the tracking URL is constructed with the hostname, thus redirecting to DESKTOP-U1EOV4J.localdomain (127.0.1.1).
For a Pseudo distributed cluster, it is safer to leave the host of JobHistoryServer to be 0.0.0.0.
Update the job history server properties in mapred-site.xml
<property>
<name>mapreduce.jobhistory.address</name>
<value>0.0.0.0:10020</value>
</property>
<property>
<name>mapreduce.jobhistory.webapp.address</name>
<value>0.0.0.0:19888</value>
</property>
and restart the JobHistoryServer.

How to increase the space for dfs on HDFS cluster

We have 4 datanode HDFS cluster ...there is large amount of space avialable on each data node of about 98gb ...but when i look at the datanode information ..
it's only using about 10gb ...
How can we make it use all the 98gb and not run out of space as indicated in image
this is the hdfs-site.xml on name node
<configuration>
<property>
<name>dfs.replication</name>
<value>2</value>
</property>
<property>
<name>dfs.name.dir</name>
<value>file:///test/hadoop/hadoopinfra/hdfs/namenode</value>
</property>
<property>
<name>hadoop.tmp.dir</name>
<value>file:///tmp/hadoop/data</value>
</property>
<property>
<name>dfs.datanode.du.reserved</name>
<value>2368709120</value>
</property>
<property>
<name>dfs.datanode.fsdataset.volume.choosing.policy</name>
<value>org.apache.hadoop.hdfs.server.datanode.fsdataset.AvailableSpaceVolumeChoosingPolicy</value>
</property>
<property>
<name>dfs.datanode.available-space-volume-choosing-policy.balanced-space-preference-fraction</name>
<value>1.0</value>
</property>
</configuration>
this is the hdfs-site.xml under data node
<configuration>
<property>
<name>dfs.replication</name>
<value>2</value>
</property>
<property>
<name>dfs.data.dir</name>
<value>file:///test/hadoop/hadoopinfra/hdfs/datanode</value>
</property>
<property>
<name>hadoop.tmp.dir</name>
<value>file:///tmp/hadoop/data</value>
</property>
<property>
<name>dfs.datanode.du.reserved</name>
<value>2368709120</value>
</property>
<property>
<name>dfs.datanode.fsdataset.volume.choosing.policy</name>
<value>org.apache.hadoop.hdfs.server.datanode.fsdataset.AvailableSpaceVolumeChoosingPolicy</value>
</property>
<property>
<name>dfs.datanode.available-space-volume-choosing-policy.balanced-space-preference-fraction</name>
<value>1.0</value>
</property>
</configuration>
the 98gb is under /test
Please let us know if we missed anything in the configuration
Look at the dfs.datanode.data.dir in the hdfs-site.xml. This property would control all the directories which can be used to store DFS blocks.
Documentation Link
So on you machines execute "df -h" that should list all the mount points which make up the 98 GB. Then in each of the mount points decide which directory can be used to store HDFS block data and add those under hdfs-site.xml comma separated for dfs.datanode.data.dir. Then retstart namenode and all data node services.
And from your edited post :
<property>
<name>dfs.data.dir</name>
<value>file:///test/hadoop/hadoopinfra/hdfs/datanode</value>
</property>
It should not be file://. It should look like :
<property>
<name>dfs.data.dir</name>
<value>/test/hadoop/hadoopinfra/hdfs/datanode</value>
</property>
Same for other properties.

GridGain No FileSystem for scheme: ggfs

everyone
I want to use GridGain in Hadoop 2.4.0
my hadoop config under that
core-site.xml
<configuration>
<property>
<name>hadoop.tmp.dir</name>
<value>/usr/hadoop-data</value>
</property>
<property>
<name>fs.trash.interval</name>
<value>1440</value>
</property>
<property>
<name>io.file.buffer.size</name>
<value>131072</value>
</property>
<property>
<name>fs.defaultFS</name>
<value>ggfs://ggfs#R</value>
</property>
<property>
<name>dfs.journalnode.edits.dir</name>
<value>/usr/hadoop-data/journal</value>
</property>
<property>
<name>hbase.zookeeper.quorum</name>
<value>r,host002,host004</value>
</property>
<property>
<name>fs.AbstractFileSystem.ggfs.impl</name>
<value>org.gridgain.grid.ggfs.hadoop.v2.GridGgfsHadoopFileSystem</value>
</property>
<property>
<name>dfs.client.block.write.replace-datanode-on-failure.policy</name>
<value>NEVER</value>
</property>
</configuration>
finish setting and start hdfs
I use
hadoop fs -ls /
ls: No FileSystem for scheme: ggfs
How should I do
Thanks
Add the followings to the core-site.xml:
<property>
<name>fs.ggfs.impl</name>
<value>org.gridgain.grid.ggfs.hadoop.v1.GridGgfsHadoopFileSystem</value>
</property>
The second version of Hadoop File System API is used rarely. The most of parts of Hadoop ecosystem works through first version of API.
And if you want to use GGFS only you don't need to start HDFS services.

Resources