Query backup from dsmc?

[ldmwndletsm]

When you run `q backup /filespace/` from dsmc on a client, is this paying any attention to the primary tape pool versus the copy pool? Is there any way to specify? Or is that moot?

So does it just query the database on the TSM server, and the server responds with what's in its table (Active only, unless -Inactive is specified), and there's simply no distinguishing what pools have the data as far as the client is concerned? The server only maintains one list (table) of the files (okay, active and inactive), but not one for the primary and a second one for the copy pool? Instead, it tracks which tapes have the files, but dsmc is simply reporting what was backed up via what the server returns, and all that happened before the copy was made (if it was), anyway, so it doesn't even know or care about the copy pool as far as the information that it returns?

 

[marclant]

So does it just query the database on the TSM server, and the server responds with what's in its table (Active only, unless -Inactive is specified), and there's simply no distinguishing what pools have the data as far as the client is concerned?

Correct.

Instead, it tracks which tapes have the files, but dsmc is simply reporting what was backed up via what the server returns, and all that happened before the copy was made (if it was), anyway, so it doesn't even know or care about the copy pool as far as the information that it returns?

It doesn't care if a copy exist or not, nor does it know if there's a copy or not.

 

[ldmwndletsm]

Thank you for the clarification.

And as far as the server is concerned, is a given object for a given node tracked separately for each pool it's on or only once, with a separate field or column to track the corresponding tapes or pools? In other words, if file system A has one million objects backed up, and there's a primary pool and copy pool, then is it just one million entries total being tracked, or does it track two million, one for each pool?