10g ASM and database.
Lets assume we got 2RAC databases: RAC1 and RAC2 sitting on one ASM diskgroup DISK1, mounted on nodes: node1, node2.
Now we want to move the RAC2 database to another cluster. So weshutdown RAC2, mount diskgroup DISK1 on another set of nodes (node11, node22), and start the RAC2. While RAC1 running from other cluster from same diskgroup.
Or to simplify, lets have single instances DB1 and DB2 running on ASM diskgroup DISK1. Now we want to move DB2 to another server - server2. Can we mount DISK1 on another server - server2, while its still mounted on server1?
Is this gonna work? Why?
Have you done this before?
If there is no Interconnect between ASM1 and storageA, and ASM2 and storageA, how would distributed locking work? There needs to be a method to prevent a process (using storageA mounted by ASM1) from overwriting changes to the same data made by a second process (using storageA mounted by ASM2).
Thus I would think that an Interconnect is essential in this case.
Yea, I think you are right, its unlikely it gonna work, I will test as soon the test env is ready. However if I manually guarantee that only one database is accessing the storage at the same time, I would like to have a option to do this.
Is ti possible to disable the asm heartbeat?