About postmigration configuration and cleanup
A successful migration results in a new disk pool for the MSDP cloud tier. If you want to use this new MSDP cloud tier server as the destination for new protection plans, policies, or duplication jobs, create a new storage unit. You must create a new storage unit for this new disk pool using the NetBackup web UI or storage API. The storage unit is not created automatically by the migration process.
Use the new storage unit as the destination for your protection plans, policies, and SLPs. You must activate any existing policies and SLPs that previously wrote to the migrated Cloud Catalyst server as the migration process disables them.
After a successful migration, you may want to clean up any obsolete objects that Cloud Catalyst created. Doing so can free up a relatively small amount of space in the cloud that is no longer needed by the MSDP cloud tier server. Cohesity recommends waiting a few days or weeks to run the cacontrol --catalog cleanupcloudcatalystobjects command until you are certain that the migration has been successful. After this command is run, there is no longer any possibility of reverting to Cloud Catalyst to access your data. This step is an optional and no functionality is affected if it is never done.
Run the following command to clean up the obsolete objects:
/usr/openv/pdde/pdcr/bin/cacontrol --catalog cleanupcloudcatalystobjects <lsuname>
During migration, the nbdecommission command asks you the following question:
Do you wish to skip migrating CloudCatalyst image sharing information? (y/n) [n]:
You can answer y to this question if you are certain that you do not use the image sharing feature in your NetBackup environment.
You should leave the default answer of n in place for all other situations or if you are unsure if your environment does not use image sharing.
You must run an additional command on the image sharing server before you can access any images that were uploaded to the cloud by Cloud Catalyst. This command should only be run if you use image sharing. Run the following command on the image sharing server:
/usr/openv/pdde/pdcr/bin/cacontrol --catalog buildcloudcatalystobjects <lsuname>
After running the cacontrol --catalog buildcloudcatalystobjects <lsuname> command, restart the NetBackup services on the image sharing server.
If backups are written directly to the Cloud Catalyst server and you have the NetBackup Accelerator option enabled on your policies, there is a special consideration for Cloud Catalyst migration. The accelerator option uses the storage server name for optimization and that storage server name changes because of migration. Therefore, the first backup job that is written to the migrated MSDP cloud tier server has no accelerator optimization. Also, for accelerator enabled multiple stream policies that write directly to the migrated MSDP cloud tier server, the deduplication rate may be zero for the first backup job. Subsequent backup jobs return to normal accelerator optimization and deduplication rates.
The migration has no effect on NetBackup Accelerator enabled policies if those policies write to MSDP and then use a duplication job to write to Cloud Catalyst.
The nbdecommission command sets MachineState to administrative pause (13) for some servers. When a server has the setting MachineState set to administrative pause (13), no jobs run, and the server may appear down.
You can display MachineState with the following command:
/usr/openv/netbackup/bin/admincmd/nbemmcmd -listhosts -display_server -machinename myserver.test.com -machinetype media -verbose
If you need to clear the administrative pause MachineState for a server, run the following command:
/usr/openv/netbackup/bin/admincmd/nbemmcmd -updatehost -machinename myserver.test.com -machinetype media -machinestateop clr_admin_pause -masterserver mymaster.test.com