On Wednesday 18 January 2006 06:40, Jo Rhett wrote:
> On Tue, Jan 17, 2006 at 10:05:58AM +0100, Kern Sibbald wrote:
> > On Tuesday 17 January 2006 05:40, Jo Rhett wrote:
> > > Interesting question. I'm getting some odd errors when trying to
> > > restore data. They don't appear to be affecting anything -- I think
> > > they happen at end of tape, but I may be wrong.
> > >
> > > Here's the Device I set up fresh. I tested this extensively with btape
> > > and it's happy:
> > >
> > > # A FreeBSD tape drive
> > > Device {
> > > Name = DDS-3
> > > Description = "DDS-3 for FreeBSD"
> > > Media Type = DDS-3
> > > Archive Device = /dev/nsa0
> > > AutomaticMount = yes; # when device opened, read it
> > > AlwaysOpen = no
> > > Offline On Unmount = no
> > > Hardware End of Medium = yes
> > > BSF at EOM = no
> > > Backward Space Record = no
> > > Fast Forward Space File = no
> > > TWO EOF = no
> > > }
> > >
> > > And here is the device I was using when I made the backups:
> > > (which I got using bextract from one of the tapes that ended this way)
> > >
> > > # A FreeBSD tape drive
> > > Device {
> > > Name = Seagate
> > > Description = "DDS-3 for FreeBSD"
> > > Media Type = DDS-3
> > > Archive Device = /dev/nsa0
> > > AutomaticMount = yes; # when device opened, read
> > > it AlwaysOpen = no
> > > Offline On Unmount = no
> > > Hardware End of Medium = no
> > > BSF at EOM = yes
> > > Backward Space Record = no
> > > Fast Forward Space File = no
> > > TWO EOF = yes
> > > LabelMedia = no;
> > > Spool Directory = /space/backups
> > > Maximum Spool Size = 15000000000
> > > }
> > >
> > > The important differences are clearly:
> > >
> > > BSF at EOM = yes
> > > TWO EOF = yes
> > >
> > > So when I'm restoring, the last (or so I believe) backup on the volume
> > > ends with the following error on 6 of the 10 volumes:
> > >
> > > 16-Jan 12:03 bextract: bextract Error: block.c:782 Read error at
> > > file:blk 8:5061 on device /dev/nsa0. ERR=Operation not permitted.
> > > 16-Jan 12:03 bextract: bextract Error: Unexpected Tape is Off-line
> > >
> > > Is this a hardware problem, (no sa0 errors are reported in the log) or
> > > a different reading of the tape data?
> >
> > It looks to me like Bacula ran off the end of the tape during the
> > bextract because either TWO EOF is not set correctly for your drive (or
> > you changed it),
>
> So that's the question. Would changing this parameter cause that error?
If you write with one set of directives and read with another, you almost
certainly will have problems. If you write and read with the same set and
btape tests pass you should not have problems.
If you wrote with TWO EOF = yes and try to restore with TWO EOF = no, you
*could* be seeing the kinds of wierd problems you reported.
> Because clearly I have changed the parameter.
>
> > and possibly due to "Hardware End of Medium = yes", which I believe is
> > not recommended (hence not supported on FreeBSD).
>
> Both of these are listed as the recommended configurations in the Bacula
> page. The former is what is now listed as "better", whereas my older
> configuration predates that comment in the docs.
You have to find the configuration that works on your system and on your tape
drive.
>
> And I guess back to the point -- would you say that this is likely the
> configuration problem for two EOF above, and can't be ignored?
--
Best regards,
Kern
(">
/\
V_V
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
