I think if there was a web based interface that would allow you to run
restores that would be a great feature. I've already coded many web based
cgi-bin utilities in perl, even a tool that lets you search for a file in
your images, but I haven't gotten to trying a web based restore GUI. This
would come in handy in our large environment where the NDMP users need to
have restore privileges on the master. That leaves us open to a security
risk because they have access to all that data that is backed on the NB
master. 

Kevin Vapiwala
Salomon Smith Barney
Vital Records Group
2nd Floor, Rutherford
Phone: 201-231-0348
Group Line: 201-231-1270

                -----Original Message-----
                From:   Peter L. Buschman [mailto:plb AT iotk DOT com]
                Sent:   Tuesday, January 22, 2002 9:49 AM
                To:     veritas-bu AT mailman.eng.auburn DOT edu
                Subject:        [Veritas-bu] Request for Comments - Unix/NT
Perl Interface to NetBackup


                Fellow NBU Users:

                I would like to pose the following question to the list:

                         "What features would you most like to see in a perl
interface to 
                NetBackup?"

                As a perl programming exercise, I have been working on an
abstract module 
                that provides read-access
                to nearly every piece of configuration and index data within
NetBackup.  I 
                have also successfully ported
                the module to Windows such that scripts that use the module
can run 
                un-modified on both Unix and NT
                platforms (so far tested only on Solaris and Windows 2000.)

                To date, I have even surprised myself at how well this
programming exercise 
                has functioned, to the point
                that I think some of you folks out there might be able to
make some use of 
                this code (under a suitable open
                source license, of course.)

                I have abstracted the following commands in their entirety,
returning their 
                output in easily-parsed perl
                hash structures (in particular, I fully parse the dreaded
bpdbjobs 
                -all_columns output, returning every single
                field.)  All functions can be passed search parameters that
take advantage 
                of perl's powerful matching syntax
                to return desired subsets of data (ideal for creating
functions that 
                monitor objects created in a certain time
                period....like for a daily job success/failure report.)

                ##   /usr/openv/netbackup/bin/goodies/bperrcode
                ##   /usr/openv/netbackup/bin/admincmd/bpclclients
                ##   /usr/openv/netbackup/bin/admincmd/bpdbjobs
                ##   /usr/openv/netbackup/bin/admincmd/bpimagelist
                ##   /usr/openv/netbackup/bin/admincmd/bpmedialist
                ##   /usr/openv/volmgr/bin/vmpool
                ##   /usr/openv/volmgr/bin/vmquery
                ##   /usr/openv/bin/admincmd/bpconfig
                ##   /usr/openv/bin/admincmd/bpgetconfig
                ##   /usr/openv/bin/admincmd/bpstulist
                ##   /usr/openv/bin/admincmd/bpcllist
                ##   /usr/openv/bin/admincmd/bpsyncinfo
                ##   /usr/openv/volmgr/bin/vmglob
                ##   /usr/openv/volmgr/bin/tpconfig

                Furthermore, version information and other details stored in
various text 
                files are accessible.

                What I would like to know is this:

                Would anyone be interested in this?  (I still have some
documentation I 
                need to write on basic usage
                since this is still just a low-level toolkit, but I have
written a few 
                example scripts that illustrate the power
                of the API.)

                If there is interest, what features would you like to see
added in the 
                future?  I am rather enjoying this as
                a fun programming project and am looking for suggestions as
to where to go 
                with it.  Since I spend my
                free time doing this, I might as well code up things that
other people find 
                useful.

                A few things that have come to mind include:

                         + A reporting framework a'la NetBackup Advanced
Reporter.
                         + Porting a number of the monitoring scripts that
are out there so 
                they are cross-platform.
                         + Adding command and control functions.
                         + An NBU configuration auditor that looks for
common problems and 
                errors.

                If there is enough interest, I will look at setting up a
mailing list for 
                cross-platform backup programming issues
                (I'd eventually like to create toolkit interfaces to Legato 
                Networker,  Tivoli Storage Manager, and others, allowing
                some application-independent tasks like media managment to
be handled 
                uniformly, but I am getting way ahead
                of myself there.)

                --PLB



                _______________________________________________
                Veritas-bu maillist  -  Veritas-bu AT mailman.eng.auburn DOT edu
                http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu