On Tue, 29 Apr 2003 17:47:14 -0700, Tom Webster
<webster AT SSDPDC.LGB.CAL.BOEING DOT COM> wrote:
>I need to be able to search for some files that may or may not have
>been on our servers. Of course, this would have been a couple of
>years ago and the user isn't real sure what the filename would have
>been.
Don't you just love users?
>I'd prefer not to have to read a couple years worth of index data back
>in using scanner. My understanding is that this should be done as a
>serial process, and with a couple hundred tapes for that area it would
>take us months.
Having to scan that amount of tapes is a real nightmare scenario.
>I've found that I can use the following to get a rough list of the
>files on a per-tape basis:
>
>scanner -c client1 -c client2 -c client3 /dev/ntmtXX -x uasm -rvn >
log.log 2>&1
>
>This generates a list, that I can search though later using grep, but
>it isn't the most ideal output structure. Adding another -v to the
>uasm command will add an additional line to the output, but there is no
>indication of the date the file was written and you have to get even the
>client name from the context of the SSID flow.
>
>Is there a way to accomplish this in a more human readable format?
>
>What I need:
>
>1. Some way to get file listings from multiple backups on multiple
> tapes. The tapes are still in the media index, but the data is
> no longer in the file index.
Without the files in the media database you have only two choices. The
first is scanner, and we've already covered that one. The second option is
to restore a client file index for that client from just after the time of
the backup. If you still have such an index on tape and in your media
database, then this is trivial using nsrck -L7 clientname, with extra
params for the date of the index - check the man page. This will merge the
index from back then with your current index. You can then use nsrinfo to
query the files in your backup. nsrck -L7 was a much underrated feature of
NetWorker 6, and could just save your bacon here. The caveat is I _think_
that you need to have been running NetWorker 6.something at or soon after
the time of your backups. I hope you were an early adopter of NetWorker 6.
>1. Speed and simplicity.
>
>2. File sizes would be nice but are not a requirement.
>
>3. Modification dates or backup dates would also be nice.
>
>4. A one-line parsable format would be very nice (something like):
>
> client::filename::size::date
>
>With some investigation into the scanner output, I may be able to
>build a parser in perl or to get something similar.
>
>Before I try to re-invent the wheel, is there an easier/better
>way to approach this?
I hope so. I know that Byron already mentioned nsrinfo in his response, if
you can get the CFI repopulated this should give you most of what you asked
for above.
Good luck.
--
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.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
|
