Namenode Issue Active and Standby

[#1]

I have a production cluster with physical server: One is active namenode other is standby name but i track the following things in my production cluster:

if i navigate the HDFS and click on quick links then it shows

nn01 is active and nn02 is standby namenode but

if i navigate to Yarn and click on quick links it shows the reverse means

nn01 is standby and nn02 is active namenode is these a problem or it happens a bigger issue in comings days....

[#2 · Accepted solution]

@ANSARI FAHEEM AHMED

From above i understood that you have enabled both Namenode HA (HDFS) and Resource Manager HA (YARN)

Now in

1) HDFS, your namenode in nn01 is active and nn02 is standby

2) YARN, your resource manager in nn01 is standby and nn02 is active

You might be confusing with the hostnames and this shouldn't be a problem.

Let me know if this is what you are looking for.

[#3]

@ANSARI FAHEEM AHMED

Sandeep Nemuri is correct, for best practice and in order to avoid confusing I request to change YARN,your resource manager to rm01,rm02

In yarn-site.xml

<property>
      <name>yarn.resourcemanager.ha.rm-ids</name>
      <value>rm01,rm02</value>
    </property>

In Ambari, YARN--> custom yarn site --> yarn.resourcemanager.ha.rm-ids = rm01,rm02

Hope this helps you.