-users@lists.sourceforge.net
Subject: Re: [Bacula-users] MANY duplicate "Max Volume jobs=X exceeded"
messages in log
Hello Adolf,
I am very happy this works for you and that this long standing
problem is finally fixed.
Many thanks for reporting this
Hello Adolf,
I am very happy this works for you and that this long standing
problem is finally fixed.
Many thanks for reporting this, noticing the connection with the
number of messages and the backup size, and re-testing and verifying
that the fix works.
Hi Kern,
A full backup ran early this morning with 184GB and there was only one
message so I can confirm that the patch has worked. Thanks very much.
Best regards,
Adolf Belka.
Sent from my Desktop Computer
On 04/10/2018 17:18, Adolf Belka wrote:
Hi Kern,
Many thanks. I have patched my
Hi Kern,
Many thanks. I have patched my current code and the next full backup of
the bacula server should occur on Sunday. Will check at that time to
confirm the fix.
Best regards,
Adolf Belka.
On 03/10/2018 17:38, Kern Sibbald wrote:
Hello Adolf and everyone,
I have now fixed the
Hello Adolf and everyone,
I have now fixed the reported problems, and the code passes all the
regression tests, so I have pushed it into the bacula git
repository. It will be available in Bacula version 9.2.2, which I
should release within a week or so. Otherwise,
Hello Aldolf and everyone,
The good news is that given your conf files, I have been able to
duplicate the problem here, and given Martin's insight, I know
roughly where the problem is coming from, though I do not yet fully
understand it as on first sight the code
Hi Kern,
llist volume=saturn-fd-Full-01 gave the following.
Automatically selected Catalog: MyCatalog
Using Catalog "MyCatalog"
MediaId: 220
VolumeName: saturn-fd-Full-01
Slot: 0
PoolId: 27
MediaType: saturnFile
MediaTypeId: 0
Hello Martin,
I think this is indeed the root of the problem.
The problem is that the Director has modified the value of VolCatStatus
while the SD
knows that it could have changed this value. It is a sort of "race"
condition where
the variable can change on both (SD and DIR) sides.
I'll
Hello Aldolf,
Can you do an "llist volume=saturn-fd-Full-01" command in bconsole?
Note: it is really "ll" (two els).
Best regards,
Kern
On 10/02/2018 03:15 PM, Adolf Belka
wrote:
Hi Kern,
I also get those
My guess is that the SD is repeated sending Update_media with the status set
to Append, even though the Director has set it to Full in has_volume_expired.
Maybe line 511 of src/stored/askdir.c is copying in the wrong direction?
bstrncpy(dcr->VolCatInfo.VolCatStatus,
Hello,
FWIW, I see the same behavior, with what I believe is a very similar
config, when running a virtual full job. The attached log excerpt is
from a virtual full job that ran as the only active job, although there
were other jobs waiting for higher priority jobs. (This job was the one
Hi Kern,
I also get those messages. I have just been living with them. Below are
the outputs from a Full job. The Full has something like 184 messages
with 183.7GB backed up. In the past I have also seen Differential jobs
with size around 2 to 3 GB giving around 3 messages so it seems to
Hello,
Hmm. 107 messages and and the volumes are 107... long. Strange
coincidence. Could you provide a copy of your bacula-sd.conf as
well as the full output including other jobs running at the time of
the 107 messages. As far as I know, there is nothing that
Hello,
wt., 2 paź 2018 o 02:25 Phil Stracchino napisał(a):
>
> So, let me see if I understand this right.
>
>
If I understand it correctly then:
> (...) you're migrating the jobs one at a time to
> tape, and then ... what, recycling the disk volumes?
He is recycling tempDisk-pool in 5 days,
On 10/1/18 7:09 PM, Stieneke, Dan wrote:
> My reason for setting "Max Volume jobs=1" is that migrations will not occur
> unless a tape is marked full, used, or error. To shorten the backup window, I
> back up to striped SSDs, and from there migrate onto tape. Setting max jobs
> to 1 ensures
1
208/423-6519
-Original Message-
From: Phil Stracchino
Sent: Monday, October 1, 2018 2:45 PM
To: bacula-users@lists.sourceforge.net
Subject: Re: [Bacula-users] MANY duplicate "Max Volume jobs=X exceeded"
messages in log
On 10/1/18 4:32 PM, Stieneke, Dan wrote:
> 64
On 10/1/18 4:32 PM, Stieneke, Dan wrote:
> 64-bit Ubuntu 18.04LTS using distro-packaged Bacula, v9.0.6.
>
>
> Backing up to 100G disk volumes, max job=1, I get 107 lines in the
> output for each volume:
>
17 matches
Mail list logo