On 10/3/2005 4:42 PM Drew Tomlinson wrote:
On 10/3/2005 12:27 PM Arno Lehmann wrote:
Hello,
On 03.10.2005 21:00, Dean Waldow wrote:
I also have have learned I have to stop bacula (likely the bacula-sd
daemon) to use either 'mt -t /dev/st0 COMMAND' or 'btape ... ' for
testing. It appears that bacula has the device in use because I
defined it already in the bacula-sd.conf file.
Just FYI: unmount does what you want, and this is documented (at
least) in the description for the console command - release.
Not actually funny, and something that should be changed. I'll see if
I can come up with some suggestions, but not immediately...
It is likely documented some where but I didn't see it near the
"Testing your Tape Drive with Bacula" section. It took me a little
while to realize that. :) That seems to take care of the resource
busy error for me. However, as Maria notes, it likely doesn't help
the rest...
Hmm. Now I wonder - you are using bextract while you've got the
daemons running? Wouldn't it be easier to just recreate the catalog
and use bacula itself?
I hadn't thought about the daemons running and creating a possible
conflict... Yes, the daemons are running although no other backup
jobs are active that use this drive. I do most of my backups to
disk. I just use the tape drive to backup my 200 GB or so of various
media files occasionally.
So yes, I could use bscan to scan them all in and then use bacula to
run the tapes through again to do the actual restore. However I
intend to restore everything, update my collection, and then run a new
backup. Thus I don't really care about recovering the database
records and felt that bextract would save some time and wear and tear
on the tapes and drive.
I will try stopping all the daemons and then see if bextract will
traverse volumes. And of course I will post my results for the next
poor soul that might be experiencing the same problems as I. :)
Stopping the daemons still did not allow me to eject the tape. However
it's an external drive and powering it off/on, waiting for it's startup
routine, and then ejecting the tape seems to do the trick.
Drew
On Oct 3, 2005, at 11:40 AM, Maria McKinley wrote:
You've probably already tried this, but I have had a problem
occasionally before of getting the device busy error when trying to
use mt-st, and just stopping the bacula daemons has gotten rid of
it. Worth a shot, anyhow, if you haven't tried already. Doesn't
answer the rest of your question, though.
-m
Arno Lehmann wrote:
Hello,
On 02.10.2005 17:50, Drew Tomlinson wrote:
I'm still stuck on this. Any attempts to eject the tape with
'mt' or 'camcontrol' result in 'device busy' errors. Thus I
assume that bextract has a hold of the device and won't let the
tape eject. Is there some config or command line option that I
need to add so bextract will eject the tape at the end of each
volume?
I'd really appreciate any thoughts, suggestions, or RTFM pointers
as I lost a couple drives and just want to put all the data
stored on these tapes back on my system. Unfortunately the
catalog was lost as well.
Unfortunately, I don't remember any information usefule for you,
and I never tried bextract like that...
Now, I assume you're in sort of an emergency. You could try the
development version - there were some changes in drive locking and
release, and hopefully they also affect bextract. Or you use your
drive's "emergency eject" feature if it has one - the drives I
know do this after holding the eject button a long time.
An easier solution might be to bextract from the volumes one at a
time, ignore the problems at tape traversal, hope everything
works, and then restore the jobs crossing tape boundaries again.
Or you set up a "full" bacula install first and bscan all the
tapes into the catalog - that might take longer, though.
Arno
Thanks,
Drew
On 9/29/2005 12:03 PM Drew Tomlinson wrote:
I'm running FreeBSD 5.4 and using bextract 1.36.3 to restore
multiple tape volumes to a drive I lost. It's giving me the
following message:
29-Sep 10:39 bextract: End of Volume at file 16 on device
/dev/nsa0, Volume
"TAPE-0003|TAPE-0004|TAPE-0005|TAPE-0006|TAPE-0010|TAPE-0011|TAPE-0012"
29-Sep 10:39 bextract: bextract Warning: Wrong Volume mounted on
device /dev/nsa0: Wanted TAPE-0004 have TAPE-0003
Mount Volume "TAPE-0004" on device /dev/nsa0 and press return
when ready:
OK, I understand it wants me to mount the next volume. However
I can not get the drive to eject the tape. It's an HP DLT-30
drive. I press the eject button on the drive but get no
response. Any attempts to use the 'mt' program from FreeBSD
result in a "device busy" error. What should I try next?
Here's the relevant part of my storage daemon conf file:
---BEGIN---
Device {
Name = "HP DLT 30"
Media Type = DLT
Archive Device = /dev/nsa0
LabelMedia = yes; # lets Bacula label
unlabeled media
AutomaticMount = yes; # when device opened, read it
AlwaysOpen = yes;
RemovableMedia = yes;
RandomAccess = no;
# From the port info:
# Due to lack of some features in the FreeBSD tape driver
# implementation you MUST add some OS dependent options to
# the bacula-sd.conf file:
Hardware End of Medium = no;
Backward Space Record = no;
Backward Space File = no;
# With 2 filemart at EOT (see man mt):
Fast Forward Space File = no;
BSF at EOM = yes;
TWO EOF = yes;
}
---END---
Any ideas?
Thanks,
Drew
-------------------------------------------------------
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ibmarch.tmpl
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
