On Tue, 1 Feb 1994 15:55:16 EST Paul Zarnowski said:
>The challenge, in my mind, is to figure out how point-in-time restores
>could be accomodated (if at all) within the ADSM backup framework, without
>severely contorting the existing framework.  I like the existing framework
>for most things, but I do see a real need for point-in-time restores.
>Rather than saying "the need does not exist", I think it would be more
>useful for this group to talk about how the need might be accomodated
>within the current ADSM framework.

Unfortunately, perhaps, ADSM thinks in terms of how many copies to keep
around at most of a file.  There are also limits that say not to keep a
copy around more than N days after the original file has been noted to be
gone, but there's no guarentee that you'll have the version that was active
N days ago.

I suspect that changing ADSM to think in that frame of mind borders on
"severely consorting the existing framework." I like the notion of a
point-in-time restore (I've seen it and used it once or twice with SyBack
on our VM system), but I doubt it's a quick and easy addition to ADSM.

It might be very easy to answer, "Is a point-in-time restore possible for
this momeent:", and if it's possible it might be easy to have the client
kick the job off from the server, but it's harder to say how far back
you'll be able to go.

I hope IBM looks into this and finds these are easily solved problems.  But
it's not going to be as easy as it would be with SyBack, which keeps or
kills copies based on time, not on how many copies are in storage.

Just a thought,
Nick

--
* nlaflamm AT vma.cc.nd DOT edu
* nlaflamm AT vma.cc.nd DOT edu
* ------
* Come to the 1994 VM Workshop: June 7-11 at the University of Notre
* Dame!  User Experience talks, dialogues with IBM, and presentations
* from other vendors in an informal setting.  Contact me or sign up for
* vmwk94-l AT vma.cc.nd DOT edu for details!
***
* "40+ million users might all be politically incorrect,
* but they're the users I support."