romeyer
ADSM.ORG Member
My situation: HSM-managed GPFS filesytem with 30+ mio files
The backup task should be splitted into two or more processes to reduce incr backup duration.
For backup I can use virtual mount points to split the filesystem into some virtual parts and start the backup task with different dsm.opt files. Works great.
But if I try to migrate a file from one of the virt. mountpoints I get the message:
Migration skipped - no backup copy found. (backup req is active).
In dsm.sys the default server stanze for migration knows all virt mountpoints and all are in the domain statement. This is different from the server stanzas for backup, which know only one mountpoint.
The backup task should be splitted into two or more processes to reduce incr backup duration.
For backup I can use virtual mount points to split the filesystem into some virtual parts and start the backup task with different dsm.opt files. Works great.
But if I try to migrate a file from one of the virt. mountpoints I get the message:
Migration skipped - no backup copy found. (backup req is active).
In dsm.sys the default server stanze for migration knows all virt mountpoints and all are in the domain statement. This is different from the server stanzas for backup, which know only one mountpoint.