Hello all --
I'm not sure how many others have run into this problem, but according to
the Bacula User's Guide, there are two ways to "blank" an unformatted (fresh
from the spindle) DVD disc (in my case I'm using / testing DVD+RW discs):
1.) Not the entire disc:
# dd if=/dev/zero bs=1024 count=512 | growisofs -Z /dev/xxx=/dev/fd/0
>
2.) The entire disc:
# growisofs -Z /dev/xxx=/dev/zero
>
It could be merely the environment I'm using: Bacula 2.2.3, growisofs
7.0with the Bacula patch from Debian, running on a Debian
4.0r1 system as (and yes this may sound convoluted but it works) a Guest
Operating System as a virtual machine using Parallels Desktop 3.0 for the
Mac, running on top of Mac OS X Server 10.4.10 (the Host Operating System)
on an Intel Xserve (with Apple's internal build-to-order slot-loading
SuperDrive option). I had at first tried to get a patched version of
growisofs to work natively on Mac OS X Server but was unsuccessful (but I
would like to re-visit this so that I could remove the extra workaround I
have come up with). So what's really cool is that Parallels offers Bridged
Ethernet capabilities such that I can give my Debian Guest OS its own static
IP address on the same LAN as my Xserve (even though the Mac OS X Server
Host OS has its own separate static IP address). So I have the Bacula
Director and File Daemons running on Mac OS X Server, which communicate over
the wire (thanks to Bridged Ethernet by Parallels) with the Storage Daemon
running inside the Debian Guest OS via Parallels (just as if Debian were
running on its own dedicated hardware attached to the same subnet).
Parallels maps the Xserve slot-loading SuperDrive at a low level perfectly,
provisioning it to the Guest OS (Debian). So all of the command-line tools
that Bacula Storage Daemons leverages (mkisofs, growisofs, dvd-handler,
etc.) work quite well.
Getting back the aforementioned, what I find a bit strange however is that
partially blanking a new DVD+RW out of the box yields error for me, as in
(where /dev/cdrom is how Debian, as standard practice, see the mapped
SuperDrive on the Xserve):
Executing 'builtin_dd if=/dev/fd/0 of=/dev/cdrom obs=32k seek=0'
> 512+0 records in
> 512+0 records out
> 524288 bytes (524 kB) copied:-[ GET EVENT failed with
> SK=5h/ASC=24h/ACQ=00h]: Input/output error
> /dev/cdrom: pre-formatting blank DVD+RW...
> , 0.126177 seconds, 4.2 MB/s
> /dev/cdrom: "Current Write Speed" is 4.1x1352KBps.
> builtin_dd: 256*2KB out @ average 0.4x1352KBps
> /dev/cdrom: flushing cache
> /dev/cdrom: stopping de-icing
> :-[ STOP DE-ICING failed with SK=4h/ASC=44h/ACQ=00h]: Input/output
> error
> /dev/cdrom: writing lead-out
>
Note: the first I/O error in the output above (regarding "GET EVENT" is
something that I always see such as when following Richard Mortimer's
excellent Bacula DVD tips web
page<http://bridge.oldelvet.org.uk/bacula/BaculaDVDSetup.html>for
creating stand-alone tests with both growisofs and dvd-handler and
have
been able to ignore).
Fascinating, what does de-icing mean? For whatever reason, the de-icing
process ran into an I/O error. Even so, I tried to request the Storage
Daemon (via the Bacula Console) to back up to this disc that was not
properly de-iced, and I had subsequent fatal errors.
However, if I formatted an entire DVD+RW disc, I had no problems getting the
Storage Daemon to complete a backup job writing to the fully blanked DVD.
Question: how do DVD disc media vendors (TDK, Verbatim, et al) ship DVD
media? I presume that they do not fill the bits on their DVDs with ASCII
NULL characters (which is what happens when utilizing the /dev/zero device),
true?
I hope this might be helpful to others, though I expect that I'm probably
the only person on the planet who has such a unique configuration using
Bacula with virtual machines on Apple Xserve hardware. I'd be happy to
provide more details about this unusual configuration if there are others
who are interested (I imagine with time as virtual machine technology
increases there will be other Bacula administrators who will want the option
of running Bacula on virtual machines).
Cheers,
Hydro
P.S. A big thanks to Richard and also Dave Green in New Zealand who has also
been climbing the Bacula DVD learning curve.
-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
