Failover of STU's

Greetings!
In our enviroment we a have 2 media servers in a cluster.
It looks like this:
Node1
Node2
Node-database (virtual name of cluster)

From "HighAvaliability Guide"->"About configuring virtual storage units in a NetBackup cluster" we managed to make 2 storage units with same "Node-database" as media but with different diskpools
STU1: DiskPool:DataDomain1 Media:"Node-database"
STU2: DiskPool:DataDomain2 Media:"Node-database"

Each node of cluster sees different DataDomains (so 1 media server for example is in storage server Group1 and another in Group2 and the virtual name resides on both storage servers)
It looks like this:
Node1 sees DataDomain1
Node2 sees DataDomain2


What we try to achieve:
If at first we have Node1 as active node we want the backup to go to DataDomain1
and if we have Node2 as active node we want the backup to go to DataDomain2 (without administrator interference (without switching different storage units in policies)

We tried these steps:
1. When Node1 was active we tried to do standard "etc/hosts" backup from hostname "Node-database" using STU1 , it succeeded

2. When Node2 was active we tried to backup from "Node-database" using STU2 and it failed with error 12 saying this:
08.11.2017 13:16:33 - Error bpbrm (pid=29118) from client Node-database: ERR - Could not open file /usr/openv/netbackup/db/config/shm/Node-database_1510135890 to get shared memory information. Errno = 2: No such file or directory
08.11.2017 13:16:35 - Error bptm (pid=29158) media manager terminated by parent process
08.11.2017 13:16:44 - Info bpbkar (pid=23458) done. status: 12: file open failed
08.11.2017 13:16:44 - end writing; write time: 0:05:05
file open failed  (12)