Different disks for NN and DN

I have 9 nodes and I have started to install HDP 2.3 using Ambari 2.1.0.

Objective : Use ONE disk for namenode, metadata etc. and rest of the disks for storing the HDFS data blocks

Node-1 : 7 disks(1 for root, opt etc., 6 empty)

df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/mapper/vg00-root
                       16G  2.8G   13G  19% /
tmpfs                  24G     0   24G   0% /dev/shm
/dev/sdb1             194M   58M  127M  32% /boot
/dev/mapper/vg00-home
                       16G   11M   15G   1% /home
/dev/mapper/vg00-nsr   16G  2.4G   13G  16% /nsr
/dev/mapper/vg00-opt   16G  260M   15G   2% /opt
/dev/mapper/vg00-itm  434M  191M  222M  47% /opt/IBM/ITM
/dev/mapper/vg00-tmp   16G   70M   15G   1% /tmp
/dev/mapper/vg00-usr   16G  2.0G   13G  14% /usr
/dev/mapper/vg00-usr_local
                      248M  231M  4.4M  99% /usr/local
/dev/mapper/vg00-var   16G  4.6G   11G  31% /var
/dev/mapper/vg00-tq   3.0G  974M  1.9G  34% /opt/teamquest
AFS                   8.6G     0  8.6G   0% /afs
/dev/sdc1             551G  198M  522G   1% /opt/dev/sdc
/dev/sdd1             551G  198M  522G   1% /opt/dev/sdd
/dev/sde1             551G  198M  522G   1% /opt/dev/sde
/dev/sdf1             551G  198M  522G   1% /opt/dev/sdf
/dev/sdg1             551G  198M  522G   1% /opt/dev/sdg
/dev/sdh1             551G  198M  522G   1% /opt/dev/sdh

Node-2 to Node-9 : 8 disks(1 for root, opt etc., 7 empty)

df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/mapper/vg00-root
                       16G  405M   15G   3% /
tmpfs                  24G     0   24G   0% /dev/shm
/dev/sda1             194M   58M  126M  32% /boot
/dev/mapper/vg00-home
                       16G   11M   15G   1% /home
/dev/mapper/vg00-nsr   16G  2.4G   13G  17% /nsr
/dev/mapper/vg00-opt   16G   35M   15G   1% /opt
/dev/mapper/vg00-itm  434M  191M  221M  47% /opt/IBM/ITM
/dev/mapper/vg00-tmp   16G   70M   15G   1% /tmp
/dev/mapper/vg00-usr   16G  1.9G   14G  13% /usr
/dev/mapper/vg00-usr_local
                      248M   11M  226M   5% /usr/local
/dev/mapper/vg00-var   16G  1.8G   14G  12% /var
/dev/mapper/vg00-tq   3.0G  946M  1.9G  33% /opt/teamquest
AFS                   8.6G     0  8.6G   0% /afs
/dev/sdb1             551G  215M  522G   1% /opt/dev/sdb
/dev/sdc1             551G  328M  522G   1% /opt/dev/sdc
/dev/sdd1             551G  215M  522G   1% /opt/dev/sdd
/dev/sde1             551G  198M  522G   1% /opt/dev/sde
/dev/sdf1             551G  198M  522G   1% /opt/dev/sdf
/dev/sdg1             551G  327M  522G   1% /opt/dev/sdg
/dev/sdh1             551G  243M  522G   1% /opt/dev/sdh

In ‘Assign Masters’

Node-1 Namenode, Zookeeper server
Node-2 SNN, RM, Zookeeper server, History Server
Node-3 WebHCat Server, HiveServer2, Hive Metastore, HBase Master, Oozie Server, Zookeeper server
Node-4 Kafka, Accumulo Master etc.
Node-5 Falcon, Knox etc.

In ‘Assign Slaves and Clients’, nothing selected for Node-1, rest 8 having uniform clients, NodeManager, Regionserver etc.

I AM STUCK IN the ‘Customize Services’, FOR EXAMPLE :

Under ‘HDFS/Namenode directories’, the defaults are :

/nsr/hadoop/hdfs/namenode
/opt/hadoop/hdfs/namenode
/opt/IBM/ITM/hadoop/hdfs/namenode
/tmp/hadoop/hdfs/namenode
/usr/hadoop/hdfs/namenode
/usr/local/hadoop/hdfs/namenode
/var/hadoop/hdfs/namenode
/opt/teamquest/hadoop/hdfs/namenode
/dev/isilon/hadoop/hdfs/namenode
/opt/dev/sdc/hadoop/hdfs/namenode
/opt/dev/sdd/hadoop/hdfs/namenode
/opt/dev/sde/hadoop/hdfs/namenode
/opt/dev/sdf/hadoop/hdfs/namenode
/opt/dev/sdg/hadoop/hdfs/namenode
/opt/dev/sdh/hadoop/hdfs/namenode

Under ‘DataNode directories’, the defaults are :

/nsr/hadoop/hdfs/data
/opt/hadoop/hdfs/data
/opt/IBM/ITM/hadoop/hdfs/data
/tmp/hadoop/hdfs/data
/usr/hadoop/hdfs/data
/usr/local/hadoop/hdfs/data
/var/hadoop/hdfs/data
/opt/teamquest/hadoop/hdfs/data
/dev/isilon/hadoop/hdfs/data
/opt/dev/sdb/hadoop/hdfs/data
/opt/dev/sdc/hadoop/hdfs/data
/opt/dev/sdd/hadoop/hdfs/data
/opt/dev/sde/hadoop/hdfs/data
/opt/dev/sdf/hadoop/hdfs/data
/opt/dev/sdg/hadoop/hdfs/data
/opt/dev/sdh/hadoop/hdfs/data

To achieve the *Objective I mentioned in the beginning, what values shall I put and in which all places?

I thought of replacing ‘DataNode directories’ defaults with
/opt/dev/sdb
/opt/dev/sdc
/opt/dev/sdd
/opt/dev/sde
/opt/dev/sdf
/opt/dev/sdg
/opt/dev/sdh