Urgent..please help me...2 data nodes goes down.

[#1]

i have added dfs.namenode.acls.enable =true in hdfs..i have 4 data node cluster with HDP 2.1.2..i restart the services

[#2]

@sankar rao can you add here what you see in the log file?

[#3]

@Sunile Manjee

i new this hadoop env..can please assist me where clould i found log file?

[#4]

@sankar rao

Please log to server where datanodes are down

cd /var/log/hadoop/hdfs
ll -ltr

there you can find the file like this hadoop-hdfs-datanode-(hostname).log

tail -300f hadoop-hdfs-datanode-(hostname).log 

and paste the output here for analysing the issue.

[#5]

i found hadoop-hdfs-namenode-(hostname).log but i con't datanode...i found this in cd /var/opt/teradata/log/hadoop directory in my system...

[#6]

@sankar rao

check for Hadoop Log Dir Prefix property location In Ambari-->HDFS-->>Advance Hadoop env. Location pointing to that property, there you can find log files.

You need to check on datanode server.

Link

[#7]

@SBandaru

it there like this .../var/opt/teradata/log/hadoop

while i go to cd //var/opt/teradata/log/hadoop ..i couldn't find any datanode..i can find namenode,zkfc,journalnode..etc but i con't see data node..

[#8]

true is spelled wrong

[#9 · Accepted solution]

Hi..tq for your time guys..i just restart the datanodes..now all nodes are up..the is resolved..but one issue(alert) left that is block health issue( correcpted blocks 108,under replicated blokcs: 500) not resolved...should i need to run hadoop fs -fsck command?

[#10]

Yes you have to run fsk command

https://community.hortonworks.com/articles/4427/fix-under-replicated-blocks-in-hdfs-manually.html