On Monday 18 March 2013 12:17:16 Michael Leone wrote:
> OK, I guess I am using the wrong terms. Here's what I want:
> 
> I want to make an exact duplicate of a tape. Not a "clone" in the 
> technical Networker sense, but a 2nd tape, with all the same data on the 
> tape. A bit-for-bit copy, same vol id, same EVERYTHING. (that's what I 
> meant by "clone")
> 
> I then want to be able to use this tape, to do CFI recovers ("nsrck -L7") 
> from, rather than the original tape, which we will keep offsite, safely. 
> 
> I want to keep the duplicate tape here onsite, and do my recovers with it. 
> What I am hoping for is that when I do a recover of that sort, NW will see 
> the mounted, local tape as the source to do said recovers, and then just 
> use it. Then, when I am done doing all the recovers I need, I can 
> physically destroy the tape, without changing anything in the NW media 
> database - so NW will think it still exists .. and it does, but now only 
> offsite. NW won't care, as long as the media matches what NW thinks it is 
> (and being a bit-for-bit duplicate, it *should* match perfectly)
> 
> Possible? If so, how?

Not using Networker.
You can do it on a *ux machine using dd repeatedly on the nonrewinding
devices but it can be tricky as Networker is using different blocksizes for
label(s) and data.
The question is: What is wrong with a logical copy?
Pros:
- Networker will not get confused if he happens to see both tapes at the same 
time
- Easy to create. No wizardy needed.
- No issues with varying capacity on tapes due to different write speeds.

Cons:
- None

Best
Dag