> I am not sure whether your procedure will also work for expired save
> sets. I would do it with this sequence:
>
> - scan the media (scanner -m) to get the SS info into the MDB.
> - set the browse- and the retention time to a future timestamp.
> - scan the media (scanner -i) to get the CFI information or use
> "nsrck -L7 -t time" assuming the the NW server 'knows' this media
> already, especially if the index backups reside on a different volume.
>
> This will work for sure. Unfortunately, if you must use scanner, you
> must run the program twice.
You must scan twice *if* the tapes are not already in the media database,
or you don't know the sequence of the savesets (i.e., the "fragflags"
field) when they span multiple tapes. If you do know which tape is first
in the sequence, you can go straight to "scanner -i".
Now, in my case, I do know the order, because I had previously done the 2
scan method:
(heavily snipped)
>mminfo -avot -q
"client=OldClient,savetime>08/30/2007,savetime<09/15/2007" -r
client,volume,name,savetime(22),nsavetime,ssflags,fragflags
client volume name date time save time ssflags
fl
OldClient Tape3 H:\Users 8/31/2007 6:18:28 PM 1188598708 vF tb
OldClient Tape2 H:\Users 8/31/2007 6:18:28 PM 1188598708 vF mb
OldClient Tape1 H:\Users 8/31/2007 6:18:28 PM 1188598708 vF hb
So, from this, I already know I have to "scenner -i" on Tape1 first,
followed by Tape2, and then by Tape3. ("Fragflags=hb" is head, or first,
tape in the sequence; "mb" is middle; "tb" is tail, or last). When you
have multiple "mbs" then it becomes problematic, but you can still figure
it out; it's just a whole lot more tedious ...
You only need to "scanner -m", followed by "scanner -i", when the tapes
are not in the media database at all. In this case, the tapes are already
in there. The contents of the tapes are not in the index, so that's why
you need to do "scanner -i".
(I realize you know that, but for the benefit of others reading, who may
find themselves in the same situation)
|
