Collectives™ on Stack Overflow
Find centralized, trusted content and collaborate around the technologies you use most.
Learn more about Collectives
Teams
Q&A for work
Connect and share knowledge within a single location that is structured and easy to search.
Learn more about Teams
root# bin/hadoop fs -mkdir t
mkdir: org.apache.hadoop.hdfs.server.namenode.SafeModeException: Cannot create directory /user/root/t. Name node is in safe mode.
not able to create anything in hdfs
I did
root# bin/hadoop fs -safemode leave
But showing
safemode: Unknown command
what is the problem?
Solution: http://unmeshasreeveni.blogspot.com/2014/04/name-node-is-in-safe-mode-how-to-leave.html?m=1
–
–
In order to forcefully let the namenode leave safemode, following command should be executed:
bin/hadoop dfsadmin -safemode leave
You are getting Unknown command error for your command as -safemode isn't a sub-command for hadoop fs, but it is of hadoop dfsadmin.
Also after the above command, I would suggest you to once run hadoop fsck so that any inconsistencies crept in the hdfs might be sorted out.
Update:
Use hdfs command instead of hadoop command for newer distributions. The hadoop command is being deprecated:
hdfs dfsadmin -safemode leave
hadoop dfsadmin has been deprecated and so is hadoop fs command, all hdfs related tasks are being moved to a separate command hdfs.
–
–
–
–
–
safe mode on means (HDFS is in READ only mode)
safe mode off means (HDFS is in Writeable and readable mode)
In Hadoop 2.6.0, we can check the status of name node with help of the below commands:
TO CHECK THE name node status
$ hdfs dfsadmin -safemode get
TO ENTER IN SAFE MODE:
$ hdfs dfsadmin -safemode enter
TO LEAVE SAFE mode
~$ hdfs dfsadmin -safemode leave
If you use Hadoop version 2.6.1 above, while the command works, it complains that its depreciated. I actually could not use the hadoop dfsadmin -safemode leave because I was running Hadoop in a Docker container and that command magically fails when run in the container, so what I did was this. I checked doc and found dfs.safemode.threshold.pct in documentation that says
Specifies the percentage of blocks that should satisfy the minimal
replication requirement defined by dfs.replication.min. Values less
than or equal to 0 mean not to wait for any particular percentage of
blocks before exiting safemode. Values greater than 1 will make safe
mode permanent.
so I changed the hdfs-site.xml into the following (In older Hadoop versions, apparently you need to do it in hdfs-default.xml:
<configuration>
<property>
<name>dfs.safemode.threshold.pct</name>
<value>0</value>
</property>
</configuration>
If it is still in safemode ,then one of the reason would be not enough space in your node, you can check your node disk usage using :
df -h
if root partition is full, delete files or add space in your root partition and retry first step.
Namenode enters into safemode when there is shortage of memory. As a result the HDFS becomes readable only. That means one can not create any additional directory or file in the HDFS. To come out of the safemode, the following command is used:
hadoop dfsadmin -safemode leave
If you are using cloudera manager:
go to >>Actions>>Leave Safemode
But it doesn't always solve the problem. The complete solution lies in making some space in the memory. Use the following command to check your memory usage.
free -m
If you are using cloudera, you can also check if the HDFS is showing some signs of bad health. It probably must be showing some memory issue related to the namenode. Allot more memory by following the options available. I am not sure what commands to use for the same if you are not using cloudera manager but there must be a way. Hope it helps! :)
