The errors below are just straight I/O errors from the underlying
device. Nothing to do with Bacula. So check if those two tapes are
bad. Check if that drive needs cleaning. Run drive diagnostics on that
drive. See if you can reproduce the problem with just, say, tar of some
files to a kno
Try changing your Maximum Network Buffer size in your bacula-sd config.
Something like
Maximum Network Buffer Size = 262144 #65536
Maximum block size = 262144
Keep in mind that this will make your sd unable to read previous
backups, IIRC.
Search archives for this parameter, e.g.
http://old.
http://www.cyberciti.biz/faq/linux-traffic-shaping-using-tc-to-control-http-traffic
Regards,
--
Alex Chekholko ch...@genomics.upenn.edu
--
This SF.net email is sponsored by
Make an app they can't live without
E
>
There's no easy way to do that. You'll want to generate a list of the files
you don't want to back up, and feed that list to the Exclude directive of the
job.
You can either use the previous backup jobs to generate the list, or do a find
on those workstat
Block Size" for LTO-4 performance) but
didn't test any restores from the tapes written with the old block
size. And eventually they were recycled.
Regards,
--
Alex Chekholko ch...@genomics.upenn.edu
--
This SF
15 years'
> time assuming the tapes hadn't become corrupt. I understand that Bacula
> uses its own tape format and that over time, this has changed.
>
The format is documented and the utilities required to read/write it
are Free Software. So it's equivalent to, say, GN
works ?
Your description is a bit confusing.
Are these tapes all in the same pool? What is the retention period?
Read up on the retention periods for files, jobs, volumes.
Do you do an "update slots" after you remove the tape? Bacula
rives on their own
> controllers even if it is supported on the RAID controller.
>
> But as you say, that support is a feature of the card and is not a
> normal feature of RAID controllers.
When you go into the RAID card's BIOS interface at boot, does it
detect and show all of device
s BIOS interface at boot, does it detect
all of devices you have physically connected to it?
--
Alex Chekholko ch...@genomics.upenn.edu
--
ThinkGeek and WIRED's GeekDad team up for the Ultimate
GeekDad Father'
6-8-10.
> Do not forget to mount the drive!!!
> You have messages.
> *
You probably forgot an 'update slots'. But more generally, make sure all of
the btape tests work first (see tape testing section of the manual.
Regards,
--
Alex Chekholko ch...@genomics.upenn.edu
---
drives are IBM
and my library is IBM and I don't use any special driver, just the 'st' kernel
module or whatever its official name is.
What is your output of 'cat /proc/scsi/scsi' or check dmesg for relevant
messages or also try the 'tapeinfo' command on
.pl MyCatalog" 11-May
17:08 bac-dir JobId 10262:
BeforeJob: /opt/bacula/bin/dbcheck: error while loading shared
libraries: libbacsql-5.0.1.so: cannot open shared object file: No such
file or directory
The fix was as Foo suggested (my libs a
conf: Maximum Concurrent Jobs = 6
bacula-dir.conf: Maximum Concurrent Jobs = 6
bacula-fd.conf: Maximum Concurrent Jobs = 20
bacula-sd.conf: Maximum Concurrent Jobs = 20
clients.conf: maximum concurrent jobs = 6
--
Alex Ch
ay being totally manual: Bacula is configured with
bacula-sd.your.domain.com as the SD address and then each client has a local
/etc/hosts entry that points to the appropriate IP for that client.
The dir connects to the fd and tells the fd to connect to the sd. But it
pass
_not_
working with Bacula; I think there used to be many more problems "back
in the day", but now things are better. That MSL6030 looks good.
Regards,
--
Alex Chekholko ch...@genomics.upenn.edu
Hi,
We've had a couple of threads about this before, but I just tried it now with
Bacula 5.0.2 and just wanted to report my technique. Comments are welcome.
Here's what I did:
1) Wrote a new pool definition:
Pool {
Name = ArchiveMay2010
Pool Type = Backup
Volume Retention = 11 months
On Mon, 15 Mar 2010 13:37:48 +1100
Norberto Meijome wrote:
> >
> >
> > Any time you add/remove tapes from the tape library, you need to do
> > 'update slots' and 'label barcodes' in bacula.
> >
> > Does that not behave as you expe
just insert new tapes and let bacula automatically label them using
> the barcode read from the tape.
>
> Is this possible, and if so, how ?
>
Any time you add/remove tapes from the tape library, you need to do
'update slots' and 'label barcodes' in b
iles_limit | 8192 |
+--+---+
1 row in set (0.03 sec)
I don't think I've ever had more than 6 jobs running concurrently.
Regards,
--
Alex Chekholko ch...@pcbi.upenn.edu
--
Download Inte
ded as you expect?
If you're worried about a lower-level problem, use 'mt -f /dev/nstX
status' and 'tapeinfo -f /dev/sgX' to see the status of the drive.
Then you can try something simple like
ommand
>
> Any ideas what I did wrong? Any hints where I can start debugging?
> Because since Baculas mtx-changer is using mtx, I think I am not able to
> use the second drive.
>
Check dmesg for relevant messages and do something like
ls -lah /dev/*st*
to see if they're th
: SKSV=no
> READ ELEMENT STATUS Command Failed
How about "tapeinfo"?
# tapeinfo -f /dev/sg10
Product Type: Medium Changer
Vendor ID: 'IBM '
Product ID: '3576-MTL'
Revision: '570G'
Attached Changer: No
SerialNumber: '01312306_LLA
r.
You can search for my name and "maximum block size". But there doesn't seem
to be any benefit to much larger block size.
--
Alex Chekholko ch...@pcbi.upenn.edu
--
This SF.Net email is sponsor
g like this:
Maximum Block Size = 262144
Maximum Network Buffer Size = 262144
Maximum File Size = 5G
Keep in mind that tapes written with different block sizes will be
incompatible, I think. So you won't be able to read your old backups
unless you r
ard and slows things
> down even further. 1.5MB/sec :-(
>
Doesn't that tell you all you need to know? Knowing this rate: 1.5MB/s
and the total amount of data in the full backup will tell you the total
time, no?
--
Alex C
sr/sbin/bregex
/usr/sbin/bscan
/usr/sbin/bsmtp
/usr/sbin/btape
/usr/sbin/btraceback
/usr/sbin/bwild
/usr/sbin/dbcheck
Regards,
--
Alex Chekholko
--
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
acula knows
about, what pool they are in (and they must be in some pool! Scratch is best
for new volumes) and what slot bacula thinks they are in the library.
--
Alex Chekholko
--
Come build with us! The BlackBerry(R
slots (SCSI)
SpectraLogic T120 with 2 LTO-3 drives and 50 slots (SCSI)
IBM TS3310 with 3 LTO-4 drives and 116 slots (SAS)
All on Linux, without any special drivers other than what is in RHEL.
Regards,
--
Alex Chekholko
--
Hi,
Familiarize yourself with the 'mt', 'tapeinfo', and 'mtx' commands. They are
standard Linux commands for dealing with the tape drives.
Then try out the 'btape' regular and autochanger tests:
http://bacula.org/3.0.x-manuals/en/utility/utility/Volume_Utility_Tools.html#SECTION0029
I haven't done that as
I've only reused tapes from other Bacula installs. A 'purge volume' or
maybe 'detete volume' should do it, check the manual.
Regards,
--
Alex Chekholko
--
Come build wi
Hi Joe,
Before dealing with bacula, make sure you understand how to use the
commands 'mt', 'mtx', 'tapeinfo' and then make sure that they are
working correctly. Then see how 'mtx-changer' uses 'mtx' to do its
business.
E.g. try:
cat /proc/scsi/scsi (this will show you the devices, 0,1,2,..)
tape
On Fri, 28 Aug 2009 10:09:46 -0400
Dan Langille wrote:
> Just published:
>
> http://dan.langille.org/2009/08/28/what-jobs-are-on-these-tapes/
>
"what jobs are on this volume" is also
query -> 13
So I usually check 'list media jobid=X' then '
acula keeps track of it somewhere.
In my case, I re-ran all full backups after changing block size, and
luckily did not need to refer to any older backups since.
Regards,
--
Alex Chekholko
--
The NEW KODAK i700
gt; DevQ(0:2:0): 0 waiting
> DevQ(0:3:0): 0 waiting
> scsi1:0:0:0: Device is active, asserting ATN
> Recovery code sleeping
> (scsi1:A:0:0): Recovery SCB completes
> Unexpected busfree in Data-in phase, 1 SCBs aborted, PRGMCNT == 0x1f5
> >>>
new one.
>
> Can anyone tell me if this is possible?
I think you just use the 'btape' command and once there, 'label' the
tape with the same label?
Regards,
--
Alex Chekholko ch...@pcbi.upenn.edu
--
Hi all,
I wanted to follow up to a number of threads we've already had about
Bacula performance with LTO-3 and LTO-4 drives.
The lowest-level way to test the drive performance is to use a utility
provided by the drive manufacturer. For IBM drives, there is a utility
called ITDT.
Here are my
y translates this list into forum form:
http://www.backupcentral.com/phpBB2/two-way-mirrors-of-external-mailing-lists-3/bacula-25/
I haven't tried posting from there, but browsing there can be quicker than the
mailing list.
On Wed, 10 Dec 2008 18:22:55 +0100
Ralf Gross wrote:
> Alex Chekholko schrieb:
> > On Tue, 9 Dec 2008 21:02:02 +0100
> > Ralf Gross wrote:
> >
> > >
> > > > # grep Maximum /etc/bacula/bacula-*.conf
> > > > /etc/bacula/bacula-dir.conf:
mber now, I added
prefer mounted volumes = no
to my Job (or JobDefs).
Regards,
--
Alex Chekholko [EMAIL PROTECTED]
--
SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
The future of the web c
Hi Ralf,
Doesn't your config do exactly what you want?
> > Storage {
> > Name = Neo4100
> > Address =
> > SDPort = 9103
> > Password = "wiped"
> > Device = Neo4100
> > Media Type = LTO4
> > Autochanger = yes
> > Heartbeat Interval = 5min
> > Maximum Concurrent Jobs = 3
> >
se my labeled tape for the new pools?
IIRC, Bacula will refuse to label tapes that already have a Bacula
label on them.
To re-use labeled tapes, I write a quick script to load each tape with
mtx, write an EOF with mt, rewind, load next...
(This will of course destroy the data on the tape.)
Th
hosts with 700 GB of data, it would take 8
> hours to enter the "select files" phase?
Bacula looks at all the file entries in the database for that, so it
depends on the number of files and how fast your db is.
> --
> David Lee Lambert ... Software Developer
> Cell phone: +1 586
= yes
> Removable Media = yes
> Random Access = no
> Requires Mount = no
> Hardware End of Medium = no
> Fast Forward Space File = no
> BSF at EOM = yes
> Two EOF = yes
> #Of
//www.adsm.org/lists/html/Bacula-users/2008-05/msg00504.html
So I am still stuck at ~22MB/s writing to LTO-4 with the default block
size.
Regards,
--
Alex Chekholko
-
This SF.Net email is sponsored by the Moblin Your Move D
t; >> ___
> >> Bacula-users mailing list
> >> Bacula-users@lists.sourceforge.net
> >> https://lists.sourceforge.net/lists/listinfo/bacula-users
> >>
> >
> >
> >
> >
On Mon, 3 Nov 2008 16:23:10 -0600
"Jeff Falgout" <[EMAIL PROTECTED]> wrote:
> >
> > *status
> >
> >
> > Device status:
> > Autochanger "T200-1" with devices:
> > "T200-1-Drive-0" (/dev/nst0)
> > "T200-1-Drive-2" (/dev/nst1)
> > "T200-1-Drive-3" (/dev/nst2)
> > Device "T200-1-Drive-0" (/dev/
Hi all,
Just wanted to chime in about making barcode labels. I'm cheap and I
happen to have some full 8.5"x11" adhesive sheets (Avery #5455) so I
just use GNU barcode and a regular laser printer and then cut the
labels by hand with a nice big paper cutter.
for ((i=1; i<=60; i++)); do printf "PIT
ate=54154.0 KB/s
> Wrote blk_block=3, dev_blk_num=10486 VolBytes=1,676,408,832
> rate=52387.8 KB/s
Hi Robin,
~50-60 MB/s is what I see on the LTO-3 based bacula install I
inherited. I believe that's regular performance. We use disk spooli
e same)
OK, at this point, can you issue, in the console:
status dir
status sd
status client
It will tell you more.
Make sure you network settings allow the dir to initiate a connection to the fd
and then allow the fd to initiate a connection to the sd.
--
Alex Chekholko
On Nov 8, 2006, at 2:05 PM, Attila Fülöp wrote:
> A Think Tank wrote:
>> I am trying to go though the rescue procedures / preparation for
>> Solaris, SunOS 5.10 Generic sun4u sparc SUNW,Sun-Fire-V240, as
>> directed
>> by the documentation. I have a couple of errors with the command
>> /usr/
On Nov 7, 2006, at 2:15 PM, Paul Cubbage wrote:
> When I install using
> bacula-mysql-1.38.11-3i386.rpm
> bacula-mtx-1.38.11-3i386.rpm
> the config files all have buildcentos4 as the hostname and the
> address = name
>
> That hostname string is nowhere on my system before I install the rpms
> but
On Nov 3, 2006, at 6:42 AM, Gerard Sharpe wrote:
> Gerard Sharpe wrote:
>> Alex Chekholko wrote:
>>
>>> On Fri, 03 Nov 2006 01:13:32 +1100
>>> Gerard Sharpe <[EMAIL PROTECTED]> wrote:
>>>
>>>
>>>
>>>> Hi,
>>&
On Fri, 03 Nov 2006 01:13:32 +1100
Gerard Sharpe <[EMAIL PROTECTED]> wrote:
> Hi,
>
> I am a first time user of Bacula trying to get a Sun L280 DLT7000 tape
> drive (operating mode set to random) with autochanger working.
>
> I have confirmed the drive works with the btape test however am havin
On Tue, 24 Oct 2006 12:17:06 -0700
"John Felczak" <[EMAIL PROTECTED]> wrote:
> This looks quite a bit better, although still not sure why mtx keeps blanking
> out on the label on the tape in slot 8.
Probably because your barcode reader doesn't read that label. Does your
library front panel sho
On Tue, 17 Oct 2006 19:51:26 +0200 (CEST)
Jonas Björklund <[EMAIL PROTECTED]> wrote:
> Hej,
>
> On Tue, 17 Oct 2006, Jon Ingason wrote:
>
> > Does anyone have a backort for Bacula 1.38 for Debian (Sarge) ADM64?
>
> What do you mean? "backort"?
>
When I type "backort" into Google, it says "Do
David Hatcher wrote:
> [snip]
>
> I just realized those four jobs that were strangely canceled had a
> priority of 11. All the jobs but the last six have a priority of 11
> (and one job as a priority of 10, but that one finish early, in fourth
> place). The last six have priorities of 13, 13, 13,
So I figured out a few more things so far, and I'd like to update my post.
> The Unix commands mt and mtx work without problems. I also checked the
> mtx-changer section of the manual
> (http://dna.dbi.udel.edu/manuals/bacula/Autochanger_Support.html#SECTION0002915000)
> and the comman
Hi all,
This is a rather lengthy e-mail, and there may be more than one
configuration error. Please feel free to respond to only a piece of
it. Configuration files and permissions info are at the end of the e-mail.
I have been trying to get bacula to work for almost three days now; I
have read
58 matches
Mail list logo