Networker

Re: [Networker] Most efficient way to create duplicate clones?

2004-05-11 11:01:17
Subject: Re: [Networker] Most efficient way to create duplicate clones?
From: Darren Dunham <ddunham AT TAOS DOT COM>
To: NETWORKER AT LISTMAIL.TEMPLE DOT EDU
Date: Tue, 11 May 2004 08:01:04 -0700
> First, as previously suggested, I've set up two clone pools: one for LTO
> tapes and one for SDLT tapes with each specifying its respective media
> type under the preferred media type field. My plan is to clone SDLT
> tapes to SDLTs and LTO to LTO's which should greatly expedite cloning
> speed by forcing the cloning to not cross media types and hence to stay
> within the same library (faster speed).

Assuming the drives are all attached to the same machine, I see no
reason why "stay within the same library" would have anything to do with
speed.  (Of course keeping the media that same should keep things fast
because speeds should be roughly matched, but that's a different issue).

> And I can do this simultaneously
> with nary a care, but not sure how best to expedite the process of
> creating a second clone copy??? I don't mind if both cloned savesets are
> on the same type of media, I just want to keep them on separate physical
> clone volumes.

You could clone the clone, rather than the master copy.  That would
guarantee that they were on separate media.

> Now, if after launching the above command, and while it's still running,
> I then launch it again, and I have a different clone volume (vol_c2)
> available, then I'm thinking NetWorker will insist on wanting to load a
> new clone volume, in this case vol_c2. Furthermore, since it can't clone
> the data until the other job finishes, I'm thinking it will just sit
> there patiently waiting, and then when the other job completes, it will
> clone the same savesets to this other tape.

Hmm.  Hopefully.

> Assuming this seems logical, anybody know of a way that I can best set
> this script up? Not sure how it will know which volume to mark read-only
> and which appendable? Seems like some careful choreography will have to
> take place. Is there a way the script can determine all this?

I think cloning the clones might be easier.

Normally you'd want to clone an ssid, but that is no longer unique now
that you've created clones.  Instead, you'd use 'mminfo' to get a list
of all the target SSIDs, but with the cloneid used on the clone tapes.
Then feed that to nsrclone.

Or, you could run all the clones once, collect a list of the cloned
media, then run nsrclone again, giving the newly cloned volumes as the
targets.

--
Darren Dunham                                           ddunham AT taos DOT com
Senior Technical Consultant         TAOS            http://www.taos.com/
Got some Dr Pepper?                           San Francisco, CA bay area
         < This line left intentionally blank to confuse you. >

--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listmail.temple DOT edu or visit the list's Web site at
http://listmail.temple.edu/archives/networker.html where you can
also view and post messages to the list.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=