We have written a set of Perl scripts to manage our off-site rotation - these
scripts essentially mimic the media management functions provided by AlphaStor.
Generally, we aren't cloning - cloning is too taxing on the system resources,
and very expensive as well, so we're sending our original backup volumes
offsite. Therefore, we're only cloning archive data and sending the clones
offsite for the duration of the archive period. We've also observed that
Legato does a much better job of tracking original backup volumes in the mmdb
than it does with clones. We use mminfo to determine which tapes have been
written to in the last 30 days and are still in the jukebox, and then we use
nsrjb to eject the tapes. We use mmlocate to update the location field in the
media database, and nsrmm to mark the tapes as full - you don't need to mark
the tapes as suspect - the whole purpose of doing that is so the system won't
try to use them while they're offsite - updating the location field to
something other than the name of the jukebox and marking the tapes full
accomplishes that for you. That's how we get tapes off-site.
As far as getting the tapes back, our off-site courier has such services
available (we would have to provide them with a return date and such, but that
didn't work out the way that we hoped). So, we have written scripts to recall
tapes as well. We use mminfo to determine which tapes are offsite and were
last written to a specified number of days back. We also search for offsite
tapes that have expired (we use this for the minimal amount of cloning that we
do). When the tapes come back, we use mminfo to update their location and we
have another script to deposit/inventory/recycle tapes back into the jukeboxes.
You are correct in that the location of the tapes will not automagically update
when you place them in the jukebox. You need to inventory the jukebox (or the
slots that the tapes are loaded into) in order to tell NetWorker that the tapes
have been placed back in the jukebox. If you don't like doing a full inventory
of the jukebox (which can be time-intensive, but is good for the system,
especially with larger - 300+ slots -jukeboxes), you can write a script that
detects non-inventoried slots in the jukebox. Use nsrjb | grep "\-\*" to
figure out which slots need to be inventoried. However, our staff feels that
using a standard nsrjb -I to run a full-scale inventory of the jukebox is worth
the time when you are moving 15 tapes a day out of each jukebox. As far as the
problem you described with full inventories and your new tapes, if you label
the tapes when you put them in the jukebox, you will not have that problem. We
have a policy in place to label new tapes once they are placed in a jukebox -
that way we always know we have a problem if we have an unidentified tape in
the library. If you have a smaller operation, that might not be such a
problem, but we have to manage 7,000 tapes.
Overall our system has been working very well, and cheap too! (we didn't have
to spend a hundred grand on AlphaStor, and it has saved our company big
insurance dollars). If you have any further questions about any of this,
please feel free to email me back.
Brandon Jackson
Carlson Companies, Inc.
Minneapolis, Minnesota
bjackson AT carlson DOT com
-----Original Message-----
From: Ed Skolnik [mailto:eskolnik AT INTERPUBLIC DOT COM]
Sent: Monday, March 03, 2003 10:24 AM
Subject: How are folks handling offsite storage? We were sending clones
off site manually, but then the clones won't scratch (recycle) till the
primary save set does.
How are folks handling offsite storage? We were sending clones off site
manually, but then the clones won't scratch (recycle) till the primary
save set does.
How I am looking to clone the world and mark the originals as full,
suspect, and then send them offsite. Some-how I'll delete (using mminfo
and nsrmm -d -y -S) the clone savesets after 14 days so I don't double
my tape usage.
How for specific questions:
I was wondering how others perform this process.
I've been playing around with updating the location on the media
database using nsrmm, but when the volume is re-deposited in the juke
box the location doesn't get automatically updated.
How you deposit takes in your juke box, When I deposit , I have to do a
full inventory of the juke box, which causes all the new tapes (never
used by networker) to be mounted and read for a label that's not found.
Does any know what command updates and reports the remarks in the media
database?
(I looked at the mmifo man page and can't find anything)
Ed Skolnik
The Interpublic Group Of Companies, Inc.
GIS Chicago System Administrator
676 North St. Clair Street
Chicago, Illinois 60611
Phone: 312-425-6554
Fax: 312-425-6924
e-mail: eskolnik AT Interpublic DOT com
This message contains information which may be confidential and
privileged. Unless you are the intended recipient (or authorized to
receive this message for the intended recipient), you may not use, copy,
disseminate or disclose to anyone the message or any information
contained in the message. If you have received the message in error,
please advise the sender by reply e-mail, and delete the message. Thank
you very much.
--
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.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
--
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.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
|