On 18/05/2010 6:19 PM, John Drescher wrote:
> On Tue, May 18, 2010 at 12:38 AM, Craig Ringer
> wrote:
>> Hi folks
>>
>> I've been having a lot of problems with Bacula's disk volume management
>> over time. Most issues seem to stem from cases where the c
; Client jobs
>947 Increme BucketSystem.2010-05-17_23.15.00_41 is waiting on max Client
> jobs
>949 Increme FlourySystem.2010-05-17_23.15.00_43 is waiting on max Client
> jobs
>951 FullBackupCatalog.2010-05-17_23.20.00_45 is waiting execution
I'm not sure I un
On 05/05/10 14:38, Henrik Johansen wrote:
> On 05/ 5/10 08:11 AM, Craig Ringer wrote:
>> Hi folks
>>
>> I've recently been saddled with a win2k8 server, and am trying to figure
>> out how to make it play with my backup infrastructure.
>>
>> Currently I&#
er with Bacula separately to the Windows image
backups, but would prefer to avoid the duplication. As there must be
people with 2k8 servers here, I thought I'd check in and see how others
are doing it.
Ideas? Suggestions?
--
Craig Ringer
Tech-related writing:
on, and it's
not as easy to do (efficiently, with proper retention etc) as it first
seems either.
--
Craig Ringer
--
___
Bacula-users mailing list
Bacula-users@li
On 01/05/10 11:03, David Young wrote:
> Is there any way, without creating 3 sets of backup jobs for each
> client, to accomplish what I'm after?
Define three jobs from the same FileSet, one for the each rotation set.
Why do you want to do this, anyway?
--
C
ignificant hassle.
Win7 and win2k8 offer an excellent built-in image backup system that
makes proper use of VSS to get a consistent copy, so this seems a bit
obsolete. I guess for 2k8 server it'd be handy.
--
Craig Ringer
Tec
stuff, I'm beginning to feel sorry I recently bought a
Win2k8 server. Still, it was that or an Apple XServe to replace the
destroyed-by-water old XServe, and unlike Apple at least Mi
nstall appropriate development packages to provide
the missing dependencies.
--
Craig Ringer
Tech-related writing: http://soapyfrogs.blogspot.com/
--
___
Bacu
er instead and not bother
with differentials.
> The diff seems like it will only do a diff from the full not another
> diff, is that correct?
Yes, that's the point.
> I was thinking Full + Incremental for 3 months.
> Initia
n't recurse to the files
within it. That's why it only mentions the directory as being marked.
Just use "mark".
The built-in help says:
mark mark dir/file to be restored recursively in dirs
markdirmark directory name to be restored (no files)
--
Craig Ringer
--
quot;setDebug" command in the console, run your differential, and
start wading though the (huge) amount of output.
See "setdebug" and "trace" in:
http://www.bacula.org/5.0.x-manuals/en/console/console/Bacula_Console.html#SECT
ed.
How are you restoring? Using a file list and "mark" ? A "restore all" ?
If you enable detailed logging of restored files (see the manual for the
messages section), does bacula-dir *think* it's restoring them?
--
Craig Ringer
---
eekend. You only need to
schedule the backup so that it'll run on a weekend, and make sure the
retention periods are set so that an old backup will expire on *or*
*before* the weekend so it can be recycled when Bacula needs it.
--
Craig Ringer
---
eone confirm this?
I've never had a problem, and sometimes have hundreds of incrementals.
However, I use disk-based storage. On tape it might well be an issue.
--
Craig Ringer
--
Download Intel® Parallel Studio Eval
Tr
file system.
Run "df -i" to see.
--
Craig Ringer
--
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel p
unix or Mac system either due to the lack of a VSS
equivalent. (I know, you can take a full file system snapshot and mount
it as I do for my mail spool backups, but clumsy?!?)
--
Craig Ringer
--
Download Intel® Para
're written to at the time of backup.
You really can't get *any* saner backup option than this?
--
Craig Ringer
--
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, fin
a would say *why* it
wanted a volume mount when it's using an auto-label pool. It's like
going next door and saying "er, can I borrow a bucket?" when what you
want to be saying is "help, help, my house is on fire!" )
--
Craig Ringer
---
r than the maximum file size permitted by this file system".
Is 81_nios2eds_linux.tar bigger than 2GB? What file system is being
restored onto?
Are you *SURE* that /backup0/bacula-restores has the free space required
(according to "df -h /backup0/ba
t to be an encoding issue, you'll need to set the
`iocharset' mount option for mount.cifs correctly. See "man mount.cifs"
and the output of the "locale" command.
If your local encoding isn't utf-8 you won't be able to reliably back up
all possible files th
stead, it just emails you to ask you to mount a volume, which isn't
exactly helpful.
--
Craig ringer
--
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, an
ting at
"Program received signal SIGSEGV, Segmentation fault" into a reply to
this email.
That, plus the information requested above, should provide some
information on what Bacula was up to when it crashed.
--
Craig Ringer
On 8/04/2010 9:56 PM, Steve Polyack wrote:
> On 04/08/10 09:34, Craig Ringer wrote:
>> On 8/04/2010 8:03 PM, Phil Stracchino wrote:
>>> On 04/08/10 02:16, Craig Ringer wrote:
>>>> Bacula should probably work with Intel's hardware crypto out of the
>>&g
On 8/04/2010 8:03 PM, Phil Stracchino wrote:
> On 04/08/10 02:16, Craig Ringer wrote:
>> Bacula should probably work with Intel's hardware crypto out of the box.
>> If it doesn't, most likely all that'd be required would be to call:
>>
>
Phil Stracchino wrote:
> On 04/08/10 03:53, Craig Ringer wrote:
>> BTW, When I suggested that greater write concurrency would be desirable
>> and should be easier, Phil Stracchino raised some concerns about
>> concurrent writes to a file system increasing fragmentation a
a controlling shell script and will
report back when that eventually finishes.
--
Craig Ringer
--
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-
ivial change that would enable Bacula to use any builtin
hardware crypto engine supported by OpenSSL. Worth making, so that by
the time the new Intel hardware hits Bacula supports it?
--
Craig Ringer
--
Download Intel® Paralle
quite sufficient for my longer term backups. Do
you actually need to retain detailed file lists for those? Remember,
they're still in the volumes, just not in the catalog database.
--
Craig Ringer
--
Download Intel® P
>> tables and indexes the same? )
>
> There are several database upgrades since 2.0.2. You need to run the
> bacula upgrade scripts. Also I would upgrade your mysql. 5.0.27 is as
> at least as ancient as bacula-2.0.2.
Yep... and has known data-loss and data-corruption bugs. Lots o
On 7/04/2010 8:42 PM, Matija Nalis wrote:
> On Wed, Apr 07, 2010 at 06:52:40PM +0800, Craig Ringer wrote:
>> Bacula currently only uses the AES CBC cypher mode. This cypher can't be
>> effectively parallelized because block n+1 depends on block n.
>>
>> The AES E
ple may need schedule-based overrides that don't neatly correspond to
job levels.
IMHO before they could be removed there'd need to be a way to override
anything, not just pool selection, by job level. Even then it's hard to
say if people would need them for other things. I guess
ntation and others are looking into it later.
The next step is to try to spawn worker threads to encrypt chunks in
parallel. Hopefully this will be possible with OpenSSL...
--
Craig Ringer
diff --git a/bacula/src/lib/crypto.c b/bacula/src/lib/crypto.c
index fb52c25..077b8d2 100644
--- a/bacula/src
n for OpenSSH I
referenced earlier was also presented to the OpenSSL folks:
http://marc.info/?l=openssl-dev&m=120180007117054&w=2
but I can't find any response or follow-up to it.
So I guess the short answer is "p
ed another one with the same hardware crypto to restore.
This is a really, really ugly situation for disaster recovery.
( Maybe things have improved since I switched away from tape, and now
the hardware crypto is just an accelerator and you can still load your
keys for driver-based crypto instead. I
Phil Stracchino wrote:
> On 04/06/10 02:37, Craig Ringer wrote:
>> Is this insane? Or a viable approach to tackling some of the
>> complexities of faking tape backup on disk as Bacula currently tries to do?
>
> Well, just off the top of my head, the first thing that comes
do, though, think Bacula's usability could use a lot of improvement.
I'm trying to throw around some ideas about some things that might help,
based on my own experiences wrangling it to get it to do what I need.
--
Craig Ringer
---
automatically just Does The Right Thing.
OK, good to know. I was actually using this mechamism before, but
thought it selected the pool *before* making the promotion decision. In
any case I had to switch to schedule-based overrides due to another
issue with retention periods and manual backups that
use it if it still works. It's not like I can't port the
config over to whatever is required later if it is ever dropped. Thanks
for the tip.
I'm troubled by the approach taken with Bacula of "it's possible with
some scripting, so
ula will no longer automatically append the
> volume ID.
>
> LabelFormat="System${Level}-${NumVols}"
I thought variable-expansion in label formats were well and truly
deprecated in favour of the Python scripting support?
--
Craig Ringer
-
re scripting where the config system could
handle it with some extension.
The scripting capabilities are great for corner cases and oddities
where the canned system will never be flexible enough. But is this such
a case, or is it actually a pretty basic configuration? I'd say the
latter, persona
ointed at
"SystemPool" for their pool, and would auto-select the pool to use based
on job level by adding a name suffix. In particular, if the job was
promoted due to an earlier failed job it'd pick the right pool for its
ne
Password = "XXX"
Device = FileStorage_HomeDir
Media Type = File_HomeDir
Maximum Concurrent Jobs = 4
}
etc
one could write:
DiskStorage {
Name = DiskSD
Devices = "File_Archival", "File_CyrusMail", "File
7;s worked ok thus far,
though I don't much like the way that different normalizations of
unicode won't match equal under SQL_ASCII so I can't reliably search for
file names.
--
Craig Ringer
--
Join us Decem
with UTF-8 encoding.
> Most of these things are done in script files, so certain non-fatal errors
> may be overlooked.
>
> As far as I can tell, it took the above encoding command, and perhaps printed
> an error message but went ahead and created the database with a
Stephen Frost wrote:
> * Craig Ringer (cr...@postnewspapers.com.au) wrote:
>> ... so it's defaulting to SQL_ASCII, but actually supports utf-8 if your
>> systems are all in a utf-8 locale. Assuming there's some way for the
>> filed to find out the encoding
On 3/12/2009 11:09 AM, Jerome Alet wrote:
> On Thu, Dec 03, 2009 at 10:54:07AM +0800, Craig Ringer wrote:
>>
>> Anyway, it'd be nice if Bacula would convert file names to utf-8 at the
>> file daemon, using the encoding of the client, for storage in a utf-8
>> data
On 3/12/2009 11:03 AM, Tom Lane wrote:
> Craig Ringer writes:
>> It's a pity that attempting to specify an encoding other than the safe
>> one when using a non-template0 database doesn't cause the CREATE
>> DATABASE command to fail with an error.
>
> Huh?
greSQL 8.4, the above command is ignored because the default
> table copied is not template0.
It's a pity that attempting to specify an encoding other than the safe
one when using a non-template0 database doesn't cause the CREATE
DATABASE command to fail with an error.
--
Craig Ringer
la ate 1.5GB of RAM and 2GB of swap during an incremental,
so I'm a bit concerned about this.
--
Craig Ringer
--
Register Now & Save for Velocity, the Web Performance & Operations
Conference from O'Reill
active" ones.
> Is anyone else doing smth like that?
Yep. I do it the dumb way - separate filesets and physical separation of
active and inactive/archived data.
--
Craig Ringer
--
Register Now & S
Martin Simmons wrote:
>>>>>> On Thu, 16 Apr 2009 10:33:43 +0800, Craig Ringer said:
>> In the above, mtime is defined as a database `timestamp' rather than
>> stored as the raw integer value from the source host. Note that the
>> database may not store
does mean that the md5 stored in the catalog is USELESS to apps
that aren't bacula, since it's not stored as a binary value or a
conventional hex-format MD5. A pity, since a binary-format MD5 would be
easy to convert to hex on-demand for external app use, but probably not
that big a
Martin Simmons wrote:
>>>>>> On Wed, 15 Apr 2009 11:58:08 +0800, Craig Ringer said:
>> Anyway, the expanded schema:
[snip]
> Thanks, I was wondering what you had done about integers. Is integer a 32-bit
> quantity?
Yes.
> Beware that many of these stat fi
Jeff Dickens wrote:
> Is it true that one needs qt 4.3 to build bat for bacula 3.0?
Only if you want to use bat, the GUI tool. And rather than build it, you
could install it from your distro's repository or backports.
--
Crai
s probably worth putting in the documentation for the
database schema, since it doesn't require installing extra tools,
compiling things, etc.
--
Craig Ringer
--
Stay on top of everything new and different, both inside
's
help on the website re debugging director/sd issues, too.
Hopefully that'll get you a MySQL error code or something more useful
than the generic error provided.
--
Craig Ringer
--
This SF.net email is sponsor
to start bconsole.
Are you sure it's not just failing to start up / timing out, rather than
exiting when you start bconsole?
Is there anything in the logs? Have you tried running the director in an
interactive terminal?
--
Craig Ringer
--
Martin Simmons wrote:
> The schema you used for the native types would be useful to know too...
Oh, sorry, I thought I included it. Schema follows, but first:
I just tested COPY ... WITH BINARY in PostgreSQL as well.
Results:
Base64 ExpandedDiff %
Dump si
ctually running ("ps -ef | grep dir") ?
Is the director shown as listening on port 9101? ("sudo netstat -ltnp |
grep 9101)
--
Craig Ringer
--
This SF.net email is sponsored by:
High Quality Requirements
. It's the SQLite / SQLite3 database file, and
you're using MySQL.
What's the actual problem you're having that the "missing" bacula.db
appears to be causing?
--
Craig Ringer
--
This SF.net e
. It only knows what the file system knows, and
few (if any) file systems keep track of MIME types.
--
Craig Ringer
--
This SF.net email is sponsored by:
High Quality Requirements in a Collaborative Environment.
Downlo
expanded-MySQL-InnoDB:
real19m27.605s
user2m16.465s
sys 1m12.693s
b64-sqlite:
real4m54.635s
user4m27.153s
sys 0m11.061s
expanded-sqlite:
real6m11.574s
user5m38.573s
sys 0m11.101s
--
Craig Ringer
---
stat' field to native database types with the
following command in Pg:
SELECT fileid, fileindex, jobid, pathid, filenameid, markid,
(decode_stat(lstat)).*,
md5
INTO file2
FROM file;
( "decode_stat" being the C extension to PostgreSQL I posted here
earlier, that gives Postgr
ling with anything but plain old ASCII. I'd seriously
consider migrating to the fr_FR.UTF-8 locale, so you use UTF-8 encoded
strings for file/directory names on the server, programs on the server
are expecting UTF-8, etc. The "C" locale bascially says "I don't care
about enco
. What happens when you run the job?
> Perhaps the job output would help.
What would *REALLY* help would be the output of EXPLAIN run on the query
that never stops. You'd need to do that from the Pg console. I posted
instructions for someone else with a
o
with the "locale" command.
On Windows, use the "nlsinfo" command (Start->Run->cmd.exe, "nlsinfo").
--
Craig Ringer
--
This SF.net email is sponsored by:
High Quality Requirements in
in the first paragraph of the email you are
> citing ... :-)
I've been using 2.5.42b2 for weeks now, by the way, and the only issues
I've encountered have since been fixed in svn.
--
Craig Ringer
--
Th
ables SQL script really needs to `\c - bacula'
(or "\c - ${bacula_user}" ... whatever) to ensure that it creates the
tables with appropriate ownership.
--
Craig Ringer
--
This SF.net email is sponsor
used to reliably
back up whole Windows systems for versions of Windows that use junction
points. Some form of post-restore repair and recovery will be required.
--
Craig Ringer
--
This SF.net email is sponsored by:
High
ve you run any
I/O benchmarks (even a simple file copy) to see if the device is
performing OK outside Bacula?
--
Craig Ringer
* Actually, it also supports a low speed for peripherals like mice and
keyboards, but nobody cares about that.
---
and most have the value 0x20 for it (not that that should
matter).
In case it matters, the file system is:
> /dev/mapper/ACCESS-ntprofiles on /ntprofiles type ext3
> (rw,noatime,acl,user_xattr)
IOW, a normal ext3 volume on top of LVM (not that th
something to do
>> with the two previous indexing which I stopped.
>
> Perhaps. Have you run any vacuums on PostgreSQL?
In particular, if you have run VACUUM FULL without a corresponding
lstat_text');
pg_relation_size
--
754434048
(1 row)
bacula=# select pg_total_relation_size('lstat_real');
pg_relation_size
--
749903872
(1 row)
... so it'd be REALLY nice if Bacula was able to j
49566630
old photos.psd | 928456141
old photos.psd | 928456141
old photos.psd | 928456141
old photos.psd | 928456141
old photos.psd | 928456141
old photos.psd |
re
directly in a bytea field, which would be even smaller.
It'd be really nice if Bacula could store this information in a more
query-friendly way.
--
Craig Ringer
--
___
Bacula
= "File_HomeDir"
Client = access.localnet-fd
FileSet = "AccessNtProfiles"
Accurate = "yes"
}
FileSet {
Name = AccessNtProfiles
Include {
Options {
signature = MD5
compression = gzip
Craig Ringer wrote:
> Thomas Mueller wrote:
>
>> as this is a beta version, and the package is not tested
>
> Youch, I spoke too soon on the upgrade.
>
> I just updated the server running the director, or tried to, as I'd
> finished updating all the FDs with yo
Johan van Vliet wrote:
> What could be writing outside of /mnt/archive/working ? I can only find one
> reference to /tmp in the config file and that is for a restore job. Not the
> case here.
Could your bootstrap files be being written in a default location like
somewhere in /var ?
Does the sd
director-mysql but it is not going to
be installed
E: Broken packages
(Aptitude just tries to "resolve" the conflict by ignoring the request
to install bacula-director-pgsql completely)
--
Craig Ringer
-
Craig Ringer wrote:
> Frank Sweetser wrote:
>> Craig Ringer wrote:
>>> Frank Sweetser wrote:
>>>
>>>> [deleted file tracking] has been implemented in the current
>>>> development code base.
>>> Oh, I also meant to ask:
>>>
>
Frank Sweetser wrote:
> Craig Ringer wrote:
>> Frank Sweetser wrote:
>>
>>> [deleted file tracking] has been implemented in the current
>>> development code base.
>>
>> Oh, I also meant to ask:
>>
>> Does deleted file tracking require th
es didn't make it into the repository dir.
--
Craig Ringer
--
Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
easily build
Frank Sweetser wrote:
> Craig Ringer wrote:
>> I note that deleted/renamed file tracking is no longer shown on the
>> sf.net bacula projects document:
>>
>> http://bacula.svn.sourceforge.net/viewvc/bacula/trunk/bacula/projects?view=markup
>>
>>
>&g
to run 2.4 and 2.5 fds in parallel on the clients.
--
Craig Ringer
--
Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
easily
s retention
periods etc to be upgraded too, it probably makes more sense.
Item 25 can already be done with scripting, and Item 24 can already be
done by splitting the fileset up appropriately and using multiple
concurrent jobs on the same fd.
--
Craig Ringer
-
vented Linux from pre-empting
already issued I/O with higher priority new I/O requests.
--
Craig Ringer
--
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open
en
managing a complex network of machines with several different OSes,
absurd volumes of data, and clumsy users. For example, being able to
effortlessly restore the newspaper's production files after a user
accidentally deleted it on deadline morning was a lifesaver.
--
Craig Ringer
-
88 matches
Mail list logo
