Re: RSTOBJ on a source physical file question

I always discourage ALWOBJDIF(*ALL) when restoring database files and members; and since some V5R# there is the ability to specify which specific differences to allow. So if *ALL was really thought to be necessary, probably ALWOBJDIF(*OWNER *FILELVL *AUTL *PGP) is desirable instead; to be clear, *ALL does not equate with all of these available options being specified.
For database file.mbr restore, think of the restore result as either restore-new, restore-rnm, or restore-rpl. The OPTION() & MBROPT() parameters along with specified FILEMBR() combinations determine most of those actions. The ALWOBJDIF() specifications can either cause or prevent /messing up/ the restore; i.e. effecting or preventing a restore-rnm result.
When the file is known to exist already, the defaulted OPTION(*ALL) is working for the given scenario. However I prefer to explicitly specify OPTION(*OLD) to be clear that the request is *not* restore-new for the *FILE, then also clarify explicitly the member processing by specifying the MBROPT(*desired_action). The more explicit the request, the less likely to have any surprises. For example, to ensure the member restores only when that same named member does not exist, specify MBROPT(*NEW). To ensure the member restores only as a replace data operation, the restore-rpl, then specify MBROPT(*OLD).
Only the ALWOBJDIF(*ALL) enables the restore-rnm which I avoid at all cost; thus naming only those attributes necessary to enable restore. Any conflict in unspecified allowed differences results in a failed restore request.

Regards, Chuck