David - I'm not entirely sure what you are asking but this posting of mine may explain things:

https://forums.emc.com/forums/thread.jspa?threadID=126901&tstart=0

There are no limitations on where you can clone from and to, though obviously you would like to ensure the data is cloned by the most efficient path... If you are using the embedded storage node functionality of the EDL there is an environmental variable you can set which will not tie the cloning to the currently mounted tape...

Hi,David
If you use the nsrclone -J options,that means you will use the storage node(-J) to recover data,so this parameter(-J) is only use for read data to clone.So it as the read host.
But when I want to clone client's backup data to Clone Pools via different storage node(all the client is not storage node),it will read the clone storage node of NetWorkwer Server's clone storage node attribute.
So clients can not clone data to physical tape library.
If I use a script to clone data,who can i delimit the node what the clone storage node writing data?
And the VTL is not EDL,so as you know the attribute of client(only) clone storage node is ignored.

I use the -J in my script to select the storage node that the source tape will be mounted on; the target tape will then be mounted on the storage node specified in the Clone Storage Node field on the client corresponding to the storage node where the tape is now mounted.

From what it sounds like your "clients" which I am interpreting to be "dedicated storage nodes" do not have the clone storage node field set correctly - there is either nothing there and it is going down the hierarchy of rules selecting the target storage node (to storage node field of backup server?)

This takes care of selecting where your backups are cloning from and to with regards to storage node; you may also be saying that you are having problems selecting where you are cloning to in terms of media pools. You need to configure your media pools to set drive restrictions to the library for where you want that pool to remain - for example your Backups pool will be configured to only access tape drives on the VTL and your Clones pool will be configured to only access tape drives on the physical tape library.

Hi,David
I know your means,If you use the -J option in your script to select the Storage node to mount tape or virtual tape,it were just only as the "read data" roles.
And my client where product data is just only the client not "dedicated storage nodes" or storage node,so it clone the data will read the clone storage node field of networker server.

If you select the delimited media pool setting drive restrictions,there may be a question with the drivers to do the other things,this means as you know,the restricted drivers will be restricted to do something like label,write,and so on.

My networker server and client version is 7.4.3.
Thanks.

So what you are asking is that you want to use the "clone storage node" field for the client to select where the data is cloned from and too? Networker does not work like that as the data no longer belongs to the client in the sense of location...
As for restrictions of devices for pools if you leave everything open you will find things start to slowly get confused and no longer work efficiently. If you configure it correctly you should have no problems...

Hi

Until the last few years, the storage node affinity rules were not documented in customer-facing documentation. Now, they are well described in the Directing clones to a special storage node section of the NetWorker Administration Guide. Note that there are separate sections for reading the clone data and writing the clone data.

If you've already read and digested this information, apologies for providing no value.

tl