On Tue, 11 Jun 2013, Leonardo - Mandic wrote:
> On old versions never have this problem, and its same network and same
> servers of old bacula versions.
I have periodically had this problem on all versions of bacula that I have
used back to 1.38, and have never been able to identify a network p
On Mon, 14 Jan 2013, John Drescher wrote:
> I would say this is a combination of filesystem performance ( remember
> that when you backup there can be a lot of seeks that reduce
> performance) and decompression performance. Decompression is less CPU
> intensive than compression.
Ah yes, you're ri
Xeon system, with all backups done
to disk. The SD is about half a mile distant). However, if I do a restore
of a large volume of data, I get 32-35 MB/sec. Seems a little odd that it
is so asymmetrical.
Steve
--
Steve
On Thu, 19 Jul 2012, Dan Langille wrote:
> On 2012-07-15 13:48, Steve Thompson wrote:
>> On Sun, 15 Jul 2012, Thomas Lohman wrote:
>>
>>> This actually is a hardcoded "sanity" check in the code itself. Search
>>> the mailing lists from the past year.
the directories whose name begins
with "s", but it also backs up st123, which has been excluded. Presumably
I have the Options clauses incorrectly defined?
TIA,
Steve
--
Steve Thompson E-mail
On Thu, 19 Jul 2012, Dan Langille wrote:
> On 2012-07-15 13:48, Steve Thompson wrote:
>> On Sun, 15 Jul 2012, Thomas Lohman wrote:
>>
>>> This actually is a hardcoded "sanity" check in the code itself. Search
>>> the mailing lists from the past year.
On Sun, 15 Jul 2012, Thomas Lohman wrote:
> This actually is a hardcoded "sanity" check in the code itself. Search
> the mailing lists from the past year. I'm pretty sure I posted where in
> the code this was and what needed to be changed.
Excellent; thank you! I have found your post and the re
On Sat, 14 Jul 2012, Boutin, Stephen wrote:
> Try changing (or adding if you don't have it already) the heartbeat
> interval variable. I have about 160TB I'm currently backing up total &
> some of the boxes are 8-29TB jobs. Heartbeat is must, for large jobs, as
> far as I'm concerned.
Good ide
On Sat, 14 Jul 2012, Joseph Spenner wrote:
> That's insane! :)
Heh :)
> Ok, can you maybe carve it up a little? How big is the backup?
I have already carved it up just about as much as I can. I have to back up
about 6 TB in 28 million files (that change very slowly) to a remote
offsite SD.
Bacula 5.0.2, CentOS 5.8.
I have this in my job definitions:
Full Max Run Time = 29d
but still they are terminated after 6 days:
14-Jul 20:27 cbe-dir JobId 39969: Fatal error: Network error with FD
during Backup: ERR=Interrupted system call
14-Jul 20:27 cbe-dir JobId 39969: Fata
On Tue, 17 Apr 2012, Martin Simmons wrote:
> Are you sure it is related to white space? I don't see anything in the above
> FileSet that would cause it. Maybe the missing directories are part of a
> different filesystem mounted on top of the main one?
There's only one file system and no nested
ept that it
does not backup any directories (and their contents) in (say)
/mnt/toe/data1/home/foo that have white space in their names. What have I
done wrong?
Steve
--
Steve Thompson E-mail: smt
On Sat, 24 Mar 2012, James Harper wrote:
>> more than one client is available to backup the (shared) storage. If I change
>> the name of the client in the Job definition, a full backup always occurs the
>> next time a job is run. How do I avoid this?
>
> That's definitely going to confuse Bacula.
Bacula 5.0.2. For the following example job:
Job {
Name = "cbe_home_a"
JobDefs = "defjob"
Pool = Pool_cbe_home_a
Write Bootstrap = "/var/lib/bacula/cbe_home_a.bsr"
Client = clarke-fd
FileSet = "cbe_home_a"
Schedule = "Saturday3"
}
FileSet {
Name = "cbe_home_a"
In
On Wed, 11 Jan 2012, Honia A wrote:
> But when I checked the size of the database it's still really large:
>
> root@servername:/var/lib/bacula# ls -l
> -rw--- 1 bacula bacula 208285783 2012-01-10 05:23 bacula.sql
Depending on what you are backing up, that is not really all that big.
Mine is
I'm using bacula 5.0.2 on CentOS 5.5. Is there any way (or any other
version of bacula) that allows one to disable generation of the
.bconsole_history file?
-steve
--
Fulfilling the Lean Software Promise
Lean software p
On Thu, 20 Jan 2011, Martin Simmons wrote:
> This will never compress -- the "default" Options clause needs to the last
> one, but you have it as the first one.
Yes, of course you are correct; thank you. And I've even read that in the
documentation. And moving the default Options clause to the e
On Thu, 20 Jan 2011, Dan Langille wrote:
> Time for new eyes. Post the job emails.
One full backup completed. Here are the relevant definitions:
Job {
Name = "bear_data15"
JobDefs = "defjob"
Pool = Pool_bear_data15
Write Bootstrap = "/var/lib/bacula/bear_data15.bsr"
Client = bear
On Thu, 20 Jan 2011, Dan Langille wrote:
> On 1/20/2011 7:24 AM, Steve Thompson wrote:
>> On Thu, 20 Jan 2011, Martin Simmons wrote:
>>
>>> It reports "None" if there were no files in the backup or if the
>>> compression
>>> saved less than
On Thu, 20 Jan 2011, Martin Simmons wrote:
> It reports "None" if there were no files in the backup or if the compression
> saved less than 0.5%, so it doesn't necessarily mean that it wasn't attempted.
I understand that, but I have several file sets that, for a full backup
level, sometimes give
On Tue, 18 Jan 2011, Dan Langille wrote:
> On 1/18/2011 4:16 PM, Steve Thompson wrote:
>> > Whether software compression happens or not seems to be random. Anyone
>> know why this is happening?
>
> There was a discussion this week about this. Add Signature to your option
Bacula 5.0.2, CentOS 5.5, x86_64.
I reported this back in November, to no comment. I have a lot of full
backups that are reporting "Software Compression: None". Software
compression is most definitely turned on. For example, all of my fileset
definitions begin in a similar fashion to:
FileSet
On Mon, 6 Dec 2010, Josh Fisher wrote:
> On 12/5/2010 9:20 AM, Steve Thompson wrote:
>> Bacula 5.0.2. This is not a problem; just an observation.
>>
>> I do backups to disk only, using six RAID arrays for storage, totalling
>> 45TB physical disk. Originally I used six
client (about 50). Backups are compressed and TLS is used
(storage is offsite). No other changes were made: backup throughput
performance almost exactly doubled.
Steve
Steve Thompson E-mail: smt AT
On Wed, 1 Dec 2010, Henrik Johansen wrote:
> The remaining posts will follow over the next month or so.
Just a minor question from part III. You state that your storage servers
each use three Perc 6/E controllers, allowing the attachment of 9 MD1000
shelves. I believe that you can attach 6 shel
me?
Steve
--------
Steve Thompson E-mail: smt AT vgersoft DOT com
Voyager Software LLC Web: http://www DOT vgersoft DOT com
39 Smugglers Path VSW Support: support AT vgersoft DOT com
Ithaca, NY 14850
"186,282 miles per s
On Thu, 26 Aug 2010, m...@free-minds.net wrote:
> 1) do we really need to spool? We are not writing to real tapes, we have a
> filesystem as backend (ext3 over glusterfs).
I am one of those that believes spooling to be useful even when writing
backups to disk; it is obviously not in question tha
On Wed, 11 Aug 2010, John Drescher wrote:
>> Bacula 5.0.2. The documentation states that a ClientRunBeforeJob script
>> that returns a non-zero status causes the job to be cancelled. This is not
>> what appears to happen, however. Instead a fatal error is declared:
>
> Maybe the documentation shou
Bacula 5.0.2. The documentation states that a ClientRunBeforeJob script
that returns a non-zero status causes the job to be cancelled. This is not
what appears to happen, however. Instead a fatal error is declared:
11-Aug 13:30 cbe-dir JobId 686: No prior Full backup Job record found.
11-Aug 13:
On Mon, 29 Mar 2010, Roland Roberts wrote:
> It was a major upgrade reboot was part of the process. And it's been
> rebooted since then.
What does a "telnet archos.rlent.pnet 9102" give you?
Steve
--
rse a
different question.
Steve
----
Steve Thompson E-mail: smt AT vgersoft DOT com Voyager Software LLC Web:
http://www DOT vgersoft DOT com 39 Smugglers Path VSW Support: support AT
vgersoft DOT com Ithaca, NY 14850
"186,300 miles per second: i
fter changing the order?
Steve
----
Steve Thompson E-mail: smt AT vgersoft DOT com
Voyager Software LLC Web: http://www DOT vgersoft DOT com
39 Smugglers Path VSW Support: support AT vgersoft DOT c
On Fri, 27 Mar 2009, John Drescher wrote:
> Have you ever seen bacula die? I mean in 5 years of using bacula on 35
> to 50 machines I do not recall ever seeing bacula die.
Yep; storage daemon (2.4.2) dies on me about once a month. I get a file
daemon failure about once a month too, but of course
On Thu, 18 Dec 2008, Stefan Sorin Nicolin wrote:
> I am about to reconfigure a mid sized Bacula intallation. I'd like to
> rename the storage daemon meaning the "Name" directive in the Storage
> { } block. Is this asking for trouble? Right now I am a bit nervous
> because I just learned (the hard
On Thu, 4 Dec 2008, Arno Lehmann wrote:
> 04.12.2008 19:41, Steve Thompson wrote:
>> Single catalog. When the second pool was added, there were 3260 volumes in
> Foo_Pool (from Foo-0001 to Foo-3260). Everything works, but the first
>> backup that went to Bar_Pool created a volu
Bacula 2.4.2.
I have just added a second pool to an all-disk configuration and have a
question concerning automatic volume numbering. The relevant details are:
Pool {
Name = Foo_Pool
Storage = Foo_Storage
Label Format = "Foo-"
...
}
Pool {
Name = Bar_Pool
y, even if there is a fancy tool to
edit them, I will stop using bacula. No doubt there are many that will
find this view unreasonable, but I can't help that.
Steve
----
Steve Thompson E-mail: sm
On Tue, 4 Nov 2008, John Drescher wrote:
>> Converting it to an XML file would not pose the problems specified in
>> the above wiki, there are lots of tools to create/parse XML files tha
>> could be useful.
>>
> I would vote against this if I could. I mean this will make it harder
> for me to e
CREATE INDEX file_tmp_pathid_idx ON File (PathId);
CREATE INDEX file_tmp_filenameid_idx ON File (FilenameId);
before the dbcheck, which produces an enormous speedup.
Steve
----
Steve Thompson E-mail:
On Fri, 29 Feb 2008, Ryan Novosielski wrote:
> Martin Simmons wrote:
>>>>>>> On Thu, 28 Feb 2008 11:20:26 -0500 (EST), Steve Thompson said:
>>> In the output of commands such as 'list jobs', is it possible to configure
>>> bacula to display nume
In the output of commands such as 'list jobs', is it possible to configure
bacula to display numeric quantities as digits alone (no commas)? I really
find this difficult to read.
-s
-
This SF.net email is sponsored by: Micr
On Mon, 24 Dec 2007, Martin Simmons wrote:
>>>>>> On Sat, 22 Dec 2007 08:29:54 -0500 (EST), Steve Thompson said:
>> [...]
>> So what is the "Files Examined" count really telling me? The JobFiles
>> count from a 'list job' is correct, howe
Here's something interesting. Bacula 2.2.4 on both client (64-bit CentOS
4.5) and director (32-bit CentOS 4.5).
During a backup:
JobId 3516 Job asimov_data7.2007-12-20_23.00.08 is running.
Backup Job started: 20-Dec-07 23:44
Files=111,850 Bytes=1,863,801,574 Bytes/sec=15,977 Errors=0
On Sun, 16 Dec 2007, David Legg wrote:
> I'm sure this is all 'obvious' to the old hacks but is there a way to
> prevent files being written into the mount point when no drive is
> actually mounted?
Just do your backups to a subdirectory on the drive which is not
present below the mount point whe
On Wed, 5 Dec 2007, Martin Simmons wrote:
> If that is OK, then I suggest running the SD with debug level 200, which
> might give us a clue where the error occurs.
So far I have been unable to get it to fail using -d200, while it does
fail if I don't specify a debug level. Maybe there is a timi
On Wed, 5 Dec 2007, Martin Simmons wrote:
>>>>>> On Wed, 5 Dec 2007 13:46:26 -0500 (EST), Steve Thompson said:
>> They are all the same at 2.2.4. It happens even in the case where
>> bacula-dir, bacula-fd and bacula-sd are running on the same machine.
>> E
On Wed, 5 Dec 2007, Dan Langille wrote:
> My first idea: different versions of SD and FD, with one trying to use a
> command the other does not recognize.
> What version is each of: bacula-dir, bacula-fd, bacula-sd
They are all the same at 2.2.4. It happens even in the case where
bacula-dir, ba
On Wed, 5 Dec 2007, Martin Simmons wrote:
>>>>>> On Wed, 5 Dec 2007 11:36:33 -0500 (EST), Steve Thompson said:
>> I see this very often as well, and I am using disk exclusively. It also
>> happens about 40% of the time, and has done since I started with bacula at
&g
On Wed, 5 Dec 2007, [EMAIL PROTECTED] wrote:
> I am still experiencing this problem on a regular basis; not every job
> does this, but it seems a good 40% do each night.
> [...]
> 05-Dec 03:33 escabot-fd JobId 8219: Fatal error: job.c:1811 Bad response
> to Append Data command. Wanted 3000 OK dat
On Wed, 26 Sep 2007, Ross Boylan wrote:
> I've been having really slow backups (13 hours) when I backup a large
> mail spool. I've attached a run report. There are about 1.4M files
> with a compressed size of 4G. I get much better throughput (e.g.,
> 2,000KB/s vs 86KB/s for this job!) with othe
; telling me?
Steve
----
Steve Thompson E-mail: smt AT vgersoft DOT com
Voyager Software LLC Web: http://www DOT vgersoft DOT com
39 Smugglers Path VSW Support: support AT vgersoft DOT com
Ithaca, NY 14850
"186,3
however, to assist in whatever way I can, given these constraints.
Steve
--------
Steve Thompson E-mail: smt AT vgersoft DOT com
Voyager Software LLC Web: http://www DOT vgersoft DOT com
39 Smugglers Path V
re-run full backups of
all of my data (about 2 TB on this system), and then restore the whole lot
to see what I get. If I get time I will take a peek at the source.
Steve
----
Steve Thompson E-mail: smt AT
Bacula 2.0.3, director is 32-bit CentOS 4.5, clients are all CentOS 4.5,
both 32-bit and 64-bit. Backups are to disk files.
I find that I cannot do any restores:
06-Sep 13:59 dante-sd: RestoreFiles.2007-09-06_13.58.29 Error: block.c:275
Volume data
error at 0:899088562! Wanted ID: "BB02
On Thu, 28 Jun 2007, David Romerstein wrote:
> On Thu, 28 Jun 2007, Steve Thompson wrote:
>
>> 28-Jun 12:33 dante-dir: inca-10_data1.2007-06-28_12.33.36 Fatal error:
>> sql_create.c:753 Create db File record INSERT INTO File
>> (FileIndex,JobId,PathId,File
Using Bacula 2.0.3 with MySQL 4.1.20 on a CentOS 4.5 x86 director. Doing a
full backup of a new file system from a 2.0.3/CentOS 4.5/x86_64 client
gives:
28-Jun 12:33 dante-dir: No prior Full backup Job record found.
28-Jun 12:33 dante-dir: No prior or suitable Full backup found in catalog.
Doin
On Tue, 20 Feb 2007, Steve Barnes wrote:
> How about coining a new "word" RTNM (read the nice manual), or RTGM
> (read the good manual) or RTBM (read the big manual). :-)
RTFBM :)
-
Take Surveys. Earn Cash. Influence the F
Bacula 2.0.1, backing up to disk exclusively, using once device and the
default pool. I get this about 20% of the time:
12-Feb 23:11 vger-dir: Start Backup JobId 601, Job=vger_u2.2007-02-12_23.05.02
12-Feb 23:11 vger-dir: Recycled volume "Backup-0065"
12-Feb 23:11 vger-sd: vger_u2.2007-02-12_23.0
different FD, it works. Anyone seen
this?
Steve
Steve Thompson E-mail: [EMAIL PROTECTED]
Voyager Software LLC Web: http://www.vgersoft.com
39 Smugglers Path VSW Support
d for minutes
at a stretch with no I/O being performed.
Steve
--------
Steve Thompson E-mail: [EMAIL PROTECTED]
Voyager Software LLC Web: http://www.vgersoft.com
39 Smugglers Path
60 matches
Mail list logo
