These are known bugs for Oracle Automatic Storage Management (Oracle ASM) in Oracle Database 12c Release 2 (12.2).
In an extended cluster, once the storage split issue has been addressed, the administrator can issue the command
crsctl modify site <site_name> -s rejuvenate to rejuvenate the site. This command, however, does not automatically restart Oracle ASM instances or its clients, such as database instances, on the rejuvenated site.
Identify the Oracle ASM instances that should be running on this site and manually issue the
srvctl start asm or the
srvctl relocate asm command.
Oracle Database 12c Release 1 (12.1) database instances cannot access disk groups with the
COMPATIBLE.ASM attribute set to Oracle Database 12c Release 2 (12.2). If attempted, the following errors are returned:
ORA-15001: diskgroup %s does not exist or is not mounted ORA-15040: diskgroup is incomplete
Apply the patch for Bug 20760821 to the Oracle Database 12c Release 1 (12.1)
Oracle Automatic Storage Management (Oracle ASM) loses the rolling migration state if Cluster Ready Services (CRS) shuts down on all nodes and the initial version of Oracle ASM is earlier than Oracle Database 12c Release 1 (184.108.40.206).
Consider the following scenario of 4 nodes (
node4) that are at Oracle Database 11g Release 2 (220.127.116.11) and that are being upgraded to Oracle Database 12c Release 2 (18.104.22.168):
node2are upgraded to 22.214.171.124 and running.
node4are still at 126.96.36.199 and running.
Now, consider that there is an outage where all CRS stacks are down which leaves the cluster in a heterogeneous state (for example, two nodes at 188.8.131.52 and two nodes at 184.108.40.206).
To proceed with the upgrade, start nodes still at 220.127.116.11 only (for example,
node4, or both) and execute the following command on the Oracle ASM instance on
node4, or both before starting any 18.104.22.168 nodes:
ALTER SYSTEM START ROLLING MIGRATION TO '22.214.171.124'
Continue the upgrade procedure as already documented from this point forward.
Note that before executing the preceding step to bring the Oracle ASM cluster back into rolling migration, you cannot start two nodes of different versions in the cluster. If you do so, one of the Oracle ASM versions fail with either the
ORA-15163 error message.