[Bacula-users] Baculum configuration

[Robert Earl]

Okay, after an apt purge of the lighttpd baculum packages, cleaning up
after the apt purge, and a reinstall, baculum-api has responded properly on
port 9096.

I am not sure what directory it wants as "Baculum working directory for
Bacula config:". There is no default provided. It needs to be writable by
the web server. The field does not match the screenshot in 9.6.x
documentation
,
and is not mentioned there. I do not think that this should be
/opt/bacula/working.

Is this a manually created directory? Does it go under the /var/www/html
tree? What permissions and attributes should it have? Does it contain any
files other than those placed there by the API?
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

[Bacula-users] Baculum/focal upgrade

[Robert Earl]

Okay, we purged and reinstalled baculum for apache2 and saw new and
different errors. It is not our job to debug apache2, so we switched to
lighttpd. It installed and ran, but now baculum is serving its web
interface on both 9095 and 9096, and we can't seem to access the API part
there.

Is baculum supported here, or does it have another community where we can
ask these questions?
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

[Bacula-users] Upgrade to focal zorched bacula again

[Robert Earl]

Hi all!
We recently upgraded from Kubuntu 18.04.5 bionic LTS, to 20.04.3 focal LTS.
Our bacula version is 9.6.7-1 from the community PPA.
Our baculum version is 9.6.6.3 as well.

The bacula-dir, bacula-fd, and bacula-sd are all running, but bconsole will
not connect. Apache is choking on rewrite rules in baculum-api.conf.

What migration docs have been published, and where can we read them to
begin repair work on this installation?

No, we will not be upgrading to v.11.

Sincerely,
Robert
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

[Bacula-users] Catalog backup failed

[Robert Earl]

Why is bacula-fd looking for the SQL database on the client? There is no
database running there, only the fd. The database is on the director
machine.

aten-sd JobId 3813: Sending spooled attrs to the Director. Despooling 0
bytes ...
aten-sd JobId 3813: Elapsed time=00:00:01, Transfer rate=0  Bytes/second
matthew-fd JobId 3813:  Could not stat
"/opt/bacula/working/bacula.sql": ERR=No such file or directory
aten-sd JobId 3813: Ready to append to end of Volume "Vol0166"
size=2,268,545,850
aten-sd JobId 3813: Volume "Vol0166" previously written, moving to end of
data.
aten-dir JobId 3813: Using Device "FileStorage" to write.
aten-dir JobId 3813: Start Backup JobId 3813,
Job=matthew-BackupCatalog.2021-09-10_14.50.35_29
aten-dir JobId 3813: shell command: run BeforeJob "/opt/bacula/scripts/
make_catalog_backup.pl matthew-Catalog"
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

Re: [Bacula-users] Encrypted restore fails with openssl errors

[Robert Earl]

OK Bacula Pros:
So I looked into the link provided about openssl and discovered that I had
reversed the order in my .pem file, putting the public CERT first and the
private KEY second. I noticed that another client fd had not been victim to
the same error. So I regenerated the PEM for the offending fd.

My next step was to do a quick backup and restore of aten to prove it was
now decryptable. However, a funny thing happened on the way to the forum.
First I tested matthew to prove it was also decryptable with no
configuration changes. The restore job went fine, until:
aten-sd JobId 3747: Elapsed time=00:00:03, Transfer rate=466  Bytes/second
matthew-fd JobId 3747: Warning: attribs.c:91 Cannot change owner and/or
group of /tmp/restore/etc/sysconfig: ERR=Operación no permitida 133 -1
matthew-fd JobId 3747: Error: attribs.c:119 Unable to set file owner
/tmp/restore/etc/sysconfig/sshd: ERR=Operación no permitida
Which is logical, because my bacula processes run unprivileged, but highly
undesirable, because it seems to imply that any large-scale restore will
end up owned by bacula:bacula entirely, and I will need to guess the
owner/group of each file? Or for a proper restore do I need to each time
swap my configuration with a root-privileged fd service?

Second unrelated snag: a "quick backup" of my server is not in the cards,
because since the last successful Full ran on 3 August and the last
successful Incremental ran on the 5th, I've been receiving this warning:
aten-dir JobId 3750: No prior Full backup Job record found.
aten-dir JobId 3750: No prior or suitable Full backup found in catalog.
Doing FULL backup.
aten-dir JobId 3750: Start Backup JobId 3750,
Job=aten-Backup.2021-08-06_15.34.17_30
And the director goes on his merry way completely preventing me from doing
the incremental at all.
And there are plainly Full backup jobs listed in Baculum, so how can the
Director be disagreeing with my view of reality?

Sincerely,
Robert

On Fri, Aug 6, 2021 at 5:30 AM Heitor Faria  wrote:

> Greetings, Bacula User Types! Long time no see!
>
> Hello Robert!
>
> Because I am in the throes of doing many dangerous maintenance tasks on my
> server, I took the liberty of testing a few restores of critical files. I
> was unsurprised to find that they all failed.
>
> aten-sd JobId 3746: Ready to read from volume "Vol0160" on File device
> "FileStorage" (/backup).
> aten-sd JobId 3746: Forward spacing Volume "Vol0160" to addr=7999614780
> aten-sd JobId 3746: Elapsed time=00:00:01, Transfer rate=2.608 K
> Bytes/second
> aten-fd JobId 3746: Error: openssl.c:68 Encryption session provided an
> invalid symmetric key: ERR=error:0407109F:rsa
> routines:RSA_padding_check_PKCS1_type_2:pkcs decoding error
> aten-fd JobId 3746: Error: openssl.c:68 Encryption session provided an
> invalid symmetric key: ERR=error:04065072:rsa
> routines:rsa_ossl_private_decrypt:padding check failed
> aten-fd JobId 3746: Error: openssl.c:68 Encryption session provided an
> invalid symmetric key: ERR=error:0607A082:digital envelope
> routines:EVP_CIPHER_CTX_set_key_length:invalid key length
> aten-fd JobId 3746: Error: restore.c:764 Failed to initialize decryption
> context for /tmp/restore/etc/bind/bind.keys
>
> Now, the configuration docs say nothing about me needing to modify config,
> as long as I have not lost keys, zorched the whole system, etc.
>
> This guy had the same error: <
> https://stackoverflow.com/questions/39228128/cant-decrypt-rsa-data-with-open-ssl
> >
>
> The troubleshooting docs, I must remark, are wafer-thin compared to the
> complexity of this enterprise software application. I did a simple Ctrl-F
> "crypt" and found no mention at all, not even in this section
> 
> ...
> I cranked up verbosity and debugging on bacula-dir
>
> The encryption tasks are performed by the bacula-fd.
>
> and ran it in the foreground as prescribed, but there is no extra logging
> anywhere that I can find (since Bacula refuses to conform to the FHS
> Filesystem Hierarchy Standard, and I had old versions from Ubuntu's repos,
> Bacula and its disused detritus is spreadeagled all over my filesystem like
> a drunken octopus.)
>
> I don't think Bacula directory setup is related to your problem.
>
> So I must throw myself upon the mercy of the community to debug this.
> Thanks.
>
> We like you, but the openssl community might be more qualified to answer
> your question.
>
> Regards,
> --
>
> MSc Heitor Faria (Miami/USA)
> Bacula LATAM CEO
> mobile1: + 1 909 655-8971
> mobile2: + 55 61 98268-4220
> [image: linkedin icon]
> 
> [image: logo] 
> América Latina
> bacula.lat | bacula.com.br 
>
>
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

[Bacula-users] Encrypted restore fails with openssl errors

[Robert Earl]

Greetings, Bacula User Types! Long time no see!

Because I am in the throes of doing many dangerous maintenance tasks on my
server, I took the liberty of testing a few restores of critical files. I
was unsurprised to find that they all failed.

aten-sd JobId 3746: Ready to read from volume "Vol0160" on File device
"FileStorage" (/backup).
aten-sd JobId 3746: Forward spacing Volume "Vol0160" to addr=7999614780
aten-sd JobId 3746: Elapsed time=00:00:01, Transfer rate=2.608 K
Bytes/second
aten-fd JobId 3746: Error: openssl.c:68 Encryption session provided an
invalid symmetric key: ERR=error:0407109F:rsa
routines:RSA_padding_check_PKCS1_type_2:pkcs decoding error
aten-fd JobId 3746: Error: openssl.c:68 Encryption session provided an
invalid symmetric key: ERR=error:04065072:rsa
routines:rsa_ossl_private_decrypt:padding check failed
aten-fd JobId 3746: Error: openssl.c:68 Encryption session provided an
invalid symmetric key: ERR=error:0607A082:digital envelope
routines:EVP_CIPHER_CTX_set_key_length:invalid key length
aten-fd JobId 3746: Error: restore.c:764 Failed to initialize decryption
context for /tmp/restore/etc/bind/bind.keys

Now, the configuration docs say nothing about me needing to modify config,
as long as I have not lost keys, zorched the whole system, etc.
The troubleshooting docs, I must remark, are wafer-thin compared to the
complexity of this enterprise software application. I did a simple Ctrl-F
"crypt" and found no mention at all, not even in this section

...
I cranked up verbosity and debugging on bacula-dir and ran it in the
foreground as prescribed, but there is no extra logging anywhere that I can
find (since Bacula refuses to conform to the FHS Filesystem Hierarchy
Standard, and I had old versions from Ubuntu's repos, Bacula and its
disused detritus is spreadeagled all over my filesystem like a drunken
octopus.)

So I must throw myself upon the mercy of the community to debug this.
Thanks.

Robert
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

[Bacula-users] Migrating file volumes between MediaTypes

[Robert Earl]

Hi Bacula Users:
Thanks for the bconsole tip. It's working ok except once when it hung. I am
using Bacula-9.6.7 (10 December 2020) x86_64-pc-linux-gnu ubuntu 18.04.

I have a situation where I was forced to create two new MediaTypes for my
volumes called "Online" and "Offline" to correspond to the ones on HDD and
BD-RE, respectively.

The MediaType was previously "File" on all volumes.

Unfortunately this has led to unusable volumes for the time being. I have
not found a way to modify the MediaType after creation and labeling. New
volumes are coming up as "Online" alright, but the database is very
confused and the director keeps attempting to create volume files which
already exist.

Oh dear, bconsole hung again:
*mount
The defined Catalog resources are:
 1: MyCatalog
 2: isidore-Catalog
 3: athanasius-Catalog
 4: matthew-Catalog
 5: felicity-Catalog
Select Catalog resource (1-5): 1
Using Catalog "MyCatalog"
The defined Storage resources are:
 1: Online
 2: Offline
Select Storage resource (1-2): 2
^C^C^C^Z
^\

Anyway, I have scoured Baculum and Bconsole and not found any way to modify
MediaType. Would a direct SQL query be necessary? What shape would that
take?

Thanks all!
Robert
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

[Bacula-users] Deadweight volumes, how to discard?

[Robert Earl]

Hi folks,
I have a large number of abandoned volumes with no data, no associated
file, and haven't been written in years. I believe they are the obstacle to
automatic labeling of new volumes; my Maximum Volumes for this pool is 120.

Alarmingly, I just found a "Full" volume whose mtime is April 9, 2021 but
Baculum's "Last Written" record is Feb. 28, 2019.

I also have 10 old volumes that are outdated and I can purge/recycle.

I am looking at batch-processing a large number of volumes, though, so this
is unwieldy through the GUI, and I'm dense on the niceties of bconsole.
What would be the procedure for a set of mass operations like this from a
script/command line?

Thanks
Robert
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

Re: [Bacula-users] Which OS is best for bacula ?

[Robert Earl]

It appears that, even if we grant that Bacula can be installed under
/opt/bacula, the FHS is still violated by installing configuration to
/opt/bacula/etc instead of /etc/opt/bacula.

This is a dangerous deviation, because firstly, /opt/bacula cannot be
NFS-shared as-is, even among same-architecture hosts, because already we'll
have to move /opt/bacula/etc somewhere else, and symlink it. Secondly, even
in the absence of shared filesystems, the administrator now needs to record
and remember that critical OS configuration is stored under /opt, which is
supposed to be, by the FHS, a place where completely replaceable software
packages are installed.

I won't even get into /opt/bacula/working vs. /var/bacula.

It would seem that the developers of enterprise backup software would
understand the vastly different backup policies that might be applied to
admin-controlled configurations, and variable log/run data, vs. a
frequently-updated software package that is downloaded, installed, and run
unmodified from its standard location. Not to mention the disk space
allocations, the simple ability of an outside party or a new hire to find
stuff when it isn't documented, etc.

So while it's nice for the Bacula team to take a principled stand on their
special needs, I would favor the Debian team's adherence to published
standards  so that systems operate as expected for
downstream consumers of this software.

Sincerely,
Robert

On Sat, Jul 10, 2021 at 7:47 AM Gary R. Schmidt 
wrote:

> On 10/07/2021 23:12, Jose Alberto wrote:
> > This brings us to other discussions haha.
> >
> > Installation preference?
> >
> What's to discuss??  Binaries, scripts, configuration and so on go under
> /opt/bacula, messages get logged to syslogd, and if the package spits
> mail, have it spit mail somewhere sensible.
>
> > Compiled or official binaries?
> >
> After the debacle with OpenSSL - Oh, look, it was Debian who did it - I
> find it very hard to trust any of the Linux distros, too many of the
> people involved are so very, very, very smart that they refuse to listen
> to anybody who has experience in the field.
>
> So build anything critical from source.
>
> I'm not quite at the NSA level of, "Delete everything on the supplied
> disk(s), compile everything from inspected source, including the kernel,
> on a known, safe system", but I'm about one more screw-up or
> Poetterer(sp?) away from it.  (And in another decade I may have enough
> super to be able to recover, and then it's SEP.  (Of course, I may take
> permanent recovery before then, which means it concerns me even less.))
>
> Fortunately, as a known BOFH, (people used to ask me if my name was
> Simon, or why don't I have a Kiwi accent), I get to specify how our
> systems are configured, and since we use different UNIX and Linux
> systems, as well as Windows, having everything in the same place on a
> system running XYZ Linux as it does on an AIX system makes life much
> easier.
>
> Cheers,
> GaryB-)
>
>
> ___
> Bacula-users mailing list
> Bacula-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/bacula-users
>
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

Re: [Bacula-users] Baculum key missing

[Robert Earl]

The SSL certificate is no longer valid for "bacula.org". Fix your URL to
read: "https://www.bacula.org/.. ." and it should
behave better.


On Thu, May 20, 2021 at 3:29 PM Jose Alberto  wrote:

> On May 18, they renewed the SSL certificate of the page
> https://www.bacula.org/. I think that has to do with it.
>
> On Thu, May 20, 2021 at 2:56 PM Gary L. Roach 
> wrote:
>
>> Operating System: Debian GNU/Linux 10
>> KDE Plasma Version: 5.14.5
>> Qt Version: 5.11.3
>> KDE Frameworks Version: 5.54.0
>> Kernel Version: 4.19.0-16-amd64
>> OS Type: 64-bit
>> Processors: 4 × AMD FX(tm)-4350 Quad-Core Processor
>> Memory: 15.6 GiB of RAM
>> I have installed bacula with minimal trouble and have run all off the
>> tests. Everything seems to be working. I then tried to install baculum and
>> got the following:
>>
>> *root@debian:/etc/apt/trusted.gpg.d#
>>  wget -qO -
>> http://bacula.org/downloads/baculum/baculum.pub
>>  |  apt-key add -
>> baculum.key*
>> *gpg: no valid OpenPGP data found.*
>> *gpg: no valid OpenPGP data found.*
>>
>> I also tried the *wget with apt-key add -* That didn't work either. What
>> am I doing wrong?
>>
>>
>> Gary R
>>
>>
>> ___
>> Bacula-users mailing list
>> Bacula-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/bacula-users
>>
>
>
> --
> #
> #   Sistema Operativo: Debian  #
> #Caracas, Venezuela  #
> #
> ___
> Bacula-users mailing list
> Bacula-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/bacula-users
>
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

Re: [Bacula-users] Certificate verification failed for Ubuntu PPAs

[Robert Earl]

This is resolved. The Let's Encrypt TLS web server certificate which was
issued yesterday, was issued for www.bacula.org and is invalid for
bacula.org. So I manually repaired the URLs listed in the
/etc/apt/sources.list.d/ and it works fine now.

On Wed, May 19, 2021 at 7:01 PM Robert Earl  wrote:

> Hi, I just performed an apt update and received these error messages for
> bacula-9.6.7 and baculum PPAs:
>
> Ign:8 https://bacula.org/downloads/baculum/stable/ubuntu bionic InRelease
> Err:17 https://bacula.org/packages/5f1e8eefd1016/debs/9.6.7/bionic/amd64
> bionic Release
>   Certificate verification failed: The certificate is NOT trusted. The
> name in the certificate does not match the expected.  Could not handshake:
> Error in the certificate verification. [IP: 80.244.178.6 443]
> Hit:18 http://ppa.launchpad.net/kubuntu-ppa/ppa/ubuntu bionic InRelease
> Err:16 https://bacula.org/downloads/baculum/stable/ubuntu bionic Release
>   Certificate verification failed: The certificate is NOT trusted. The
> name in the certificate does not match the expected.  Could not handshake:
> Error in the certificate verification. [IP: 80.244.178.6 443]
>
>
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

[Bacula-users] Certificate verification failed for Ubuntu PPAs

[Robert Earl]

Hi, I just performed an apt update and received these error messages for
bacula-9.6.7 and baculum PPAs:

Ign:8 https://bacula.org/downloads/baculum/stable/ubuntu bionic InRelease
Err:17 https://bacula.org/packages/5f1e8eefd1016/debs/9.6.7/bionic/amd64
bionic Release
  Certificate verification failed: The certificate is NOT trusted. The name
in the certificate does not match the expected.  Could not handshake: Error
in the certificate verification. [IP: 80.244.178.6 443]
Hit:18 http://ppa.launchpad.net/kubuntu-ppa/ppa/ubuntu bionic InRelease
Err:16 https://bacula.org/downloads/baculum/stable/ubuntu bionic Release
  Certificate verification failed: The certificate is NOT trusted. The name
in the certificate does not match the expected.  Could not handshake: Error
in the certificate verification. [IP: 80.244.178.6 443]
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

[Bacula-users] Automatic labeling of volumes

[Robert Earl]

I have recently changed the media types in my Devices resources and
apparently, I cannot retroactively change this on existing volumes? So I am
again faced with creating many new volumes as backups progress.

My device is regular file storage in the /backup directory. I have followed
the instructions in "Automatic Volume Labeling" in the 9.6.7 manual, namely
"LabelMedia = yes;" in the Device resource, and LabelFormat = "Vol" in the
Pool resource, but Bacula is still pausing backups and asking me to label a
new volume for use.

Robert
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

[Bacula-users] bscanning offline volumes and the role of the bootstrap file

[Robert Earl]

Hi folks, I am working on a restore of files from 2 years ago, and it's
exposed some serious flaws in my understanding of Bacula.

A little about my setup: I have a File pool on an HDD partition where my
online backups go. Then I have a manual procedure to migrate the volumes to
BD-RE XL 100GB disc every few months.
Version: 9.6.7 (10 December 2020)

I have not been migrating the .bsr files along with the Vol files, and
that seems like a bad oversight on my part. Now I am reading in the docs
about the need for .bsr when restoring. So I have symlinked to the relevant
bootstrap files and integrated them into the migration process for the
future.

My 2019 backups are spread across 2 Blu-ray discs. I ran bscan with mostly
success, but it coredumps at the very end of the first disc with this
message:

04-May-2021 21:48:41 bscan: bscan.c:658-0 Could not update job record.
ERR=bdb.h:140 Update failed: affected_rows=0 for UPDATE Job SET
JobStatus='E',EndTime='2019-03-01
23:05:00',ClientId=1,JobBytes=354680832,ReadBytes=0,JobFiles=1,JobErrors=0,VolSessionId=7,VolSessionTime=1551383812,PoolId=1,FileSetId=0,JobTDate=1551506700,RealEndTime='2019-03-01
23:05:00',PriorJobId=0,HasBase=0,PurgedFiles=0 WHERE JobId=1920

So as a result, some jobs are restored to my catalog db, but the Full
backup I'm targeting is in a failed state, or even flagged as still
running, and zero-sized. When I explore it with Baculum's restore tool, I
find /home is empty, and I would have noticed 48GB missing during the
backup reports, or the migration. Plus, the incrementals indicate files
from my home directory do exist in those backups with the same FileSets
specified.

So is there some corruption on disc? Have I tickled a bug in the bscan
utility? Is there yet hope for restoring this backup? That's kind of the
raison d'etre of running Bacula in the first place...

Robert
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

Re: [Bacula-users] Version mismatch causing auth failure?

[Robert Earl]

Thanks, I will do these debugging steps. Yes, I have already checked the
things suggested. I downgraded back to 9.6.5 and I still have the error.

Now for a slightly different Q: How do I set command-line options for the
bacula director? My installation lacks systemd integration, which is
problematic: It never restarts on boot, and I have to start it with the
manual bespoke scripts in /opt/bacula, and I don't have the expertise or
wherewithal to wrestle with systemd long enough to create a service unit
file of my very own. Bacula should be doing that automatically in the
installation, it's not my job!

Robert

On Tue, Apr 6, 2021 at 10:50 AM Martin Simmons  wrote:

> >>>>> On Mon, 5 Apr 2021 01:27:07 -0700, Robert Earl said:
> >
> > I recently upgraded Bacula from 9.6.5 to 9.6.7 on the director host. Now
> I
> > have received an authentication error when running backups on my client
> > Fedora machine.
> >
> > aten-dir JobId 3491: Start Backup JobId
> 3491,Job=matthew-Backup.2021-04-05_01.22.01_26
> > aten-dir JobId 3491: Using Device "FileStorage" to write.
> > aten-dir JobId 3491: Fatal error: Unable to authenticate with File
> daemon at "matthew-vpn.spiritus.cox.net:9102". Possible causes:
> > Passwords or names not the same or
> > Maximum Concurrent Jobs exceeded on the FD or
> > FD networking messed up (restart daemon).
> > For help, please see:
> >
> http://www.bacula.org/rel-manual/en/problems/Bacula_Frequently_Asked_Que.html
>
> Did you check these things?
>
> If so, try running the Director and the FD with debug level 50 to see what
> messages are being sent.  E.g. restart the daemons with arguments -f -d50
> added to the normal command line.
>
>
> > aten-dir JobId 3491: Fatal error: No Job status returned from FD.
> > aten-dir JobId 3491: Error: Bacula aten-dir 9.6.7 (10Dec20):
> >   Build OS:   x86_64-pc-linux-gnu ubuntu 18.04
> >   JobId:  3491
> >   Job:matthew-Backup.2021-04-05_01.22.01_26
> >   Backup Level:   Full
> >   Client: "matthew-fd" 9.6.6 (20Sep20)
> x86_64-redhat-linux-gnu-bacula,redhat,One)
> >   FileSet:"Matthew Set" 2021-04-05 00:58:25
> >   Pool:   "File" (From Command input)
> >   Catalog:"matthew-Catalog" (From Client resource)
> >   Storage:"File" (From Command input)
> >   Scheduled time: 05-Apr-2021 01:22:01
> >   Start time: 05-Apr-2021 01:22:03
> >   End time:   05-Apr-2021 01:22:14
> >   Elapsed time:   11 secs
> >   Priority:   10
> >   FD Files Written:   0
> >   SD Files Written:   0
> >   FD Bytes Written:   0 (0 B)
> >   SD Bytes Written:   0 (0 B)
> >   Rate:   0.0 KB/s
> >   Software Compression:   None
> >   Comm Line Compression:  None
> >   Snapshot/VSS:   no
> >   Encryption: no
> >   Accurate:   no
> >   Volume name(s):
> >   Volume Session Id:  6
> >   Volume Session Time:1617609357
> >   Last Volume Bytes:  1,880,205,247 (1.880 GB)
> >   Non-fatal FD errors:1
> >   SD Errors:  0
> >   FD termination status:  Error
> >   SD termination status:  Waiting on FD
> >   Termination:*** Backup Error ***
> >
> >
> > The bizarre thing now is that the log reports file director version
> 9.6.6,
> > but the FD actually running on matthew is 9.6.7, double-checked and
> > verified against packages and command output. So would a 9.6.6 client FD
> > cause this error, and where could it be lurking?
>
> Did you upgrade the FD as well?  The job log will report the last known
> version of the FD, which will be the previous version if the authentication
> fails.
>
> __Martin
>
>
> ___
> Bacula-users mailing list
> Bacula-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/bacula-users
>
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

[Bacula-users] Version mismatch causing auth failure?

[Robert Earl]

I recently upgraded Bacula from 9.6.5 to 9.6.7 on the director host. Now I
have received an authentication error when running backups on my client
Fedora machine.

aten-dir JobId 3491: Start Backup JobId 3491,
Job=matthew-Backup.2021-04-05_01.22.01_26
aten-dir JobId 3491: Using Device "FileStorage" to write.
aten-dir JobId 3491: Fatal error: Unable to authenticate with File
daemon at "matthew-vpn.spiritus.cox.net:9102". Possible causes:
Passwords or names not the same or
Maximum Concurrent Jobs exceeded on the FD or
FD networking messed up (restart daemon).
For help, please see:
http://www.bacula.org/rel-manual/en/problems/Bacula_Frequently_Asked_Que.html
aten-dir JobId 3491: Fatal error: No Job status returned from FD.
aten-dir JobId 3491: Error: Bacula aten-dir 9.6.7 (10Dec20):
  Build OS:   x86_64-pc-linux-gnu ubuntu 18.04
  JobId:  3491
  Job:matthew-Backup
.2021-04-05_01.22.01_26
  Backup Level:   Full
  Client: "matthew-fd
" 9.6.6 (20Sep20)
x86_64-redhat-linux-gnu-bacula,redhat,One)
  FileSet:"Matthew Set
" 2021-04-05 00:58:25
  Pool:   "File "
(From Command input)
  Catalog:"matthew-Catalog" (From Client resource)
  Storage:"File
" (From Command input)
  Scheduled time: 05-Apr-2021 01:22:01
  Start time: 05-Apr-2021 01:22:03
  End time:   05-Apr-2021 01:22:14
  Elapsed time:   11 secs
  Priority:   10
  FD Files Written:   0
  SD Files Written:   0
  FD Bytes Written:   0 (0 B)
  SD Bytes Written:   0 (0 B)
  Rate:   0.0 KB/s
  Software Compression:   None
  Comm Line Compression:  None
  Snapshot/VSS:   no
  Encryption: no
  Accurate:   no
  Volume name(s):
  Volume Session Id:  6
  Volume Session Time:1617609357
  Last Volume Bytes:  1,880,205,247 (1.880 GB)
  Non-fatal FD errors:1
  SD Errors:  0
  FD termination status:  Error
  SD termination status:  Waiting on FD
  Termination:*** Backup Error ***


The bizarre thing now is that the log reports file director version 9.6.6,
but the FD actually running on matthew is 9.6.7, double-checked and
verified against packages and command output. So would a 9.6.6 client FD
cause this error, and where could it be lurking?

R
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

[Bacula-users] Fwd: bacula-fd errors on backupCatalog run

[Robert Earl]

-- Forwarded message -
From: Robert Earl 
Date: Tue, Dec 8, 2020 at 9:47 PM
Subject: bacula-fd errors on backupCatalog run
To: 


So I am unable to determine whether bacula-fd in a client-only installation
requires a database layer or not.

It seems that it does not, because the client installed properly on my
Fedora 31 laptop, and bacula-fd starts and runs correctly, even according
me a valid backup of the filesystem. However, when I go to backup the
Catalog, I receive a database error. Could not stat
"/opt/bacula/working/bacula.sql": ERR=No such file or directory.

aten-dir JobId 3392: shell command: run AfterJob
"/opt/bacula/scripts/delete_catalog_backup"
aten-dir JobId 3392: End auto prune.
aten-dir JobId 3392: No Files found to prune.
aten-dir JobId 3392: Begin pruning Files.
aten-dir JobId 3392: No Jobs found to prune.
aten-dir JobId 3392: Begin pruning Jobs older than 6 months .
aten-dir JobId 3392: Bacula aten-dir 9.6.5 (11Jun20):
  Build OS:   x86_64-pc-linux-gnu ubuntu 18.04
  JobId:  3392
  Job:matthew-BackupCatalog
<http://localhost:9095/web/job/matthew-BackupCatalog>.2020-12-08_00.11.38_09
  Backup Level:   Full
  Client: "matthew-fd
<http://localhost:9095/web/client/matthew-fd>" 9.6.6 (20Sep20)
x86_64-redhat-linux-gnu-bacula,redhat,One)
  FileSet:"Catalog
<http://localhost:9095/web/fileset/Catalog>" 2020-11-28 11:38:21
  Pool:   "File <http://localhost:9095/web/pool/File>"
(From Command input)
  Catalog:"matthew-Catalog" (From Client resource)
  Storage:"File
<http://localhost:9095/web/storage/File>" (From Command input)
  Scheduled time: 08-Dec-2020 00:11:38
  Start time: 08-Dec-2020 00:12:04
  End time:   08-Dec-2020 00:12:04
  Elapsed time:   1 sec
  Priority:   11
  FD Files Written:   0
  SD Files Written:   0
  FD Bytes Written:   0 (0 B)
  SD Bytes Written:   0 (0 B)
  Rate:   0.0 KB/s
  Software Compression:   None
  Comm Line Compression:  None
  Snapshot/VSS:   no
  Encryption: yes
  Accurate:   yes
  Volume name(s):
  Volume Session Id:  11
  Volume Session Time:160731
  Last Volume Bytes:  847,895,289 (847.8 MB)
  Non-fatal FD errors:1
  SD Errors:  0
  FD termination status:  OK
  SD termination status:  OK
  Termination:Backup OK -- with warnings
aten-sd JobId 3392: Sending spooled attrs to the Director. Despooling
0 bytes ...
aten-sd JobId 3392: Elapsed time=00:00:01, Transfer rate=0  Bytes/second
matthew-fd JobId 3392:  Could not stat
"/opt/bacula/working/bacula.sql": ERR=No such file or directory
aten-sd JobId 3392: Volume "Vol0002" previously written, moving to end of data.
aten-dir JobId 3392: Using Device "FileStorage" to write.
aten-dir JobId 3392: Start Backup JobId 3392,
Job=matthew-BackupCatalog.2020-12-08_00.11.38_09
aten-dir JobId 3392: shell command: run BeforeJob
"/opt/bacula/scripts/make_catalog_backup.pl matthew-Catalog"


If I create a zero-size file in the proper place on the client, the Catalog
backup proceeds, but is of course null. Why is it even looking for a
database??
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users