Hey, folks!
I've updated my NetBackup tool set on Backup Central. The tools that are
there now have been tested in real environments against 3.2 and
3.4. There's some really neat tools here, if I do say so myself. ;) They
include tools to: give you calender-based scheduling of full backups
(autosched.pl), completely automate the backup of your database
(backupdb.sh), automate taking tapes in and out of a library (clam), tell
you when your library is almost empty (monitorjb.pl), and many
others. I've included a complete description below. You can get this free
tool bag at
http://www.backupcentral.com/mytools.html
Here is the README.txt that comes with the tar file, which includes a
complete description of each tool:
---------------------------------------------------------------------------------------------
These tools are meant to be installed in /usr/openv/local
The commands go in ./bin, the libraries in ./lib, the config files in
./etc, and
the data files in ./data.
If you cd to /usr/openv and untar the scripts, it will put everything in
the right place.
The following tools are for use with NetBackup:
autosched.pl*
This script will create a full backup schedule for each NetBackup
class. To use it,
create a file /usr/openv/local/etc/autosched.conf, with one line for
each class.
It should be 'classname:dayofmonth,' the 'dayofmonth' being the day
you want its full
backup to run. Then you put autosched.pl in cron and run it once a
day. For each
class, you must create a daily incremental schedule, because
autosched uses that to
create the 'days' file for the Full schedule. Every day it runs, it
checks to see
if any classes are supposed to receive a full backup that day. If
so, it creates
a full backup schedule with an interval of 10 days, and uses the
days file from
the incremental backup for windows. It also uses retention number 8
(You must
edit autosched.pl to change that.). It then records that it did this in
/usr/openv/local/data/autosched.dat. Nine days later, it deletes
the schedule.
This gives the class nine days to do a full backup, but ensures that
the full
backup always occurs on the same day of the month. (You can also
specify a day
of the week in autosched.conf. If so, it will create the full on
that day of the
week, with an interval of 4 days, removing it 3 days later.)
backupdb.sh*
This script will completely automate the backup of your database.
Use backupdb.sh -x to get a complete usage statment. Essentially,
it uses the
Scratch pool to grab a fresh database backup tape every day, stops
bprd before
the nbu db backup, then runs bpbackupdb. It allows you to put the
db backup
tapes in a separate pool, and allows you to rotate the oldest tapes
out of that
pool backup into the Scratch pool. To use this script, you need a
Scratch pool
and a pool for the NBDB backups. NEAT stuff.
checkdrives.sh*
This script checks all the KNOWN servers in vm.conf for down drives and
complains about it if finds them.
autocheckdrives.sh*
Just like checkdrives, but supports TNG alerts.
clam*
Curtis' library automation menu. This automates all tape management
for libraries.
It will a) take all bpvault tapes out, (b) take all tapes out
matching certain
criteria, including date, pool, and volume names (c) take out one
tape a time
(d) import new tapes from the i/o slot of your library, reinventory
your library,
and clean drives. It automatically displays a list of libraries to
choose from,
and allows you to select which one you want to work with. The only site
customization required is to edit /usr/openv/local/lib/nbu.pl and
set the
$maxslots variable to the number of slots in your i/o port.
epoch.pl*
This script converts epoch time data to human readable time.
frozen.sh*
This script finds and complains about any frozen tapes in your
environment.
gen_excludes*
This pushes a default exclude list out to every UNIX client. See
the README
file in /usr/openv/local/etc/exclude_lists for more details.
(Uses bpgp to push the files out.)
inactive_classes*
This is to be run from cron and complains about any inactive classes.
Inactive classes cause backups not to be run without generating any
errors.
monitorjb*
This script is designed to warn you about libraries that are short
on tapes.
By default, it checks every library, and requires ten free tapes per
pool.
You can change the defaults with command line optins, which can be
optained
by giving it a -x option.
nbimport.sh*
Automates the importing of a tape volume. Without this, the import
is a two-step
process.
rsh.sh*
Uses bpgp to add the master's hostname to all Unix client's .rhosts file
or take the master's hostname out. Use it to turn on rsh
temporarily for
quick fixes, without having to turn it on all the time.
The rest of these scripts apply only if you are using bpvault.
makedbtape.sh*
Pulls a tape from the Scratch pool into the database backup pool for
BPVAULT,
since bpvault does not automatically do this.
bpvault.sh*
An example script to start bpvault as soon as backups are done.
bpvault.bcv.sh*
An example script to start bpvault as soon as backups from a certain
class are done.
bpvault.copy.sh*
The reverse of the script above, this one is an example script to
start bpvault
as soon as all backups EXCEPT for certain class type are done.
killvault.sh*
Kills bpvault gracefully, starting by issuing a haltdups.now
command, then waiting
an hour. Then it kills the bpduplicate processes and waits another
hour. Then it
kills bpvault entirely. Each stage drops off if the previous stage
was successful
at getting bpvault to stop completely.
checkdups*
Checks all vaults to see if duplications are done.
---
W. Curtis Preston
Principal Consultant for Storage Designs, your storage experts
Webmaster: http://www.backupcentral.com Phone: 760 631 7991
|
