There is no such option do downgrade HDP version out of the box. Also we do not see any HDP 2.7 version released. Also HDP 3.x and HDP 2.6 has major differences in terms of components. Any specific reason you are looking out for downgrade? If this is a freshly build cluster on HDP 3. and you want to use HDP 2.6 then better to freshly install HDP 2.6 which will be lot time and effort saving than manually fixing and downgrading all the components and configs. ... View more

@Manoj690  Are you sure that the NameNode is running on "localhost" (where you are opening the mentioned URL in the browser) ? 1. Can you specify the namenode IP Address/ Hostname in the URL instead of "localhost" ? 2. Can you also check if the NameNode is listening on port 50070 ?  (is that port opened and firewall is disabled) on NameNode host? # netstat -tnlpa | grep 50070 # service iptables stop 3. Please check if you are able to telnet NamerNode host=name & port from the machine where you are running the Browser? # telnet $NAMENODE_HOST 50070 (OR) # mc -v $NAMENODE_HOST 50070 4. Check and share the NameNode log. Usually it can be found inside the " /var/log/hadoop/hdfs/hadoop-hdfs-namenode-xxxxxxxxxxxxxxxxxx.log "     ... View more

@Manoj690  Going through your logs I can see that the Namenode is in SAFE MODE, and in this case it won't allow you to change the status of any file in the cluster including the logs. 2019-08-22 12:31:01,376 [server.Accumulo] INFO : Attempting to talk to zookeeper 2019-08-22 12:31:01,681 [server.Accumulo] INFO : ZooKeeper connected and initialized, attempting to talk to HDFS 2019-08-22 12:31:01,946 [server.Accumulo] WARN : Waiting for the NameNode to leave safemode 2019-08-22 12:31:01,946 [server.Accumulo] INFO : Backing off due to failure; current sleep period is 1.0 seconds 2019-08-22 12:31:02,950 [server.Accumulo] WARN : Waiting for the NameNode to leave safemode 2019-08-22 12:31:02,950 [server.Accumulo] INFO : Backing off due to failure; current sleep period is 2.0 seconds 2019-08-22 12:31:04,954 [server.Accumulo] WARN : Waiting for the NameNode to leave safemode   To resolve the issue can you do the following as hdfs user $ hdfs dfsadmin -safemode get Safe mode is OFF The above is the desired output but if you get ON then proceed like below First backup your FS Image & Edits $ hdfs dfsadmin -saveNamespace   Then exit the safemode $ hdfs dfsadmin -safemode leave Once successful then revalidate $ hdfs dfsadmin -safemode get   This time it should be off and you can now successfully restart the failed services from Ambari everything should succeed HTH ... View more

@Manoj690  Not the same output as mine. Please notice that,  In your case that file is being owned by "root" user.  Not the "hdfs" user.  So please try this:     # chown hdfs:hadoop /hadoop/hdfs/namenode/current/VERSION (OR recursively) # chown -R hdfs:hadoop /hadoop/hdfs/       ... View more