In the meanwhile I also tried to create a separate storage in the director in
analogy to a restore storage device with AutoSelect=no. This also does not work
as the volumes I want to truncate are then not found to belong to this storage
device.
> On 19. Apr 2023, at 11:43, Justin Case <jus7inc...@gmail.com> wrote:
>
> Hi all
>
> I declared a second device in the disk autochanger, but the same problem
> occurs when trying to truncate volumes while a backup job is running.
> I also have set MaximumConcurrentJobs to 5, so this shouldn’t be the problem.
> It seems that the second device which I declared in the autochanger is not
> used for the truncating:
>
> truncate pool=pool1 storage=storage1
> Automatically selected Catalog: MyCatalog
> Using Catalog "MyCatalog"
> Connecting ...
> 3908 Error reserving volume: vol_mgr.c:418 Cannot free Volume
> “pool1-vol-0599", because it is reserved by someone else.
>
> Error truncating Volume “pool11-vol-0700"
> 3908 Error reserving volume: vol_mgr.c:418 Cannot free Volume
> “pool1-vol-0599", because it is reserved by someone else.
>
> Error truncating Volume “pool1-vol-0703”
>
>
> This is how I declared the second device in the disk autochanger (dev1 and
> dev2 are declared identical):
>
> director config file:
>
> Storage {
> Name = “storage1"
> Device = "changer1"
> MediaType = “file1"
> MaximumConcurrentJobs = 5
> }
>
> SD config file:
>
> Autochanger {
> Name = “changer1"
> Device = “restore1"
> Device = “dev1"
> Device = “dev2"
> ChangerDevice = "/dev/null"
> ChangerCommand = "/dev/null"
> }
>
> Device {
> Name = “dev1"
> MediaType = “file1"
> DeviceType = "File"
> ArchiveDevice = "/mnt/folder1"
> RemovableMedia = no
> RandomAccess = yes
> AutomaticMount = yes
> LabelMedia = yes
> AlwaysOpen = no
> Autochanger = yes
> MaximumConcurrentJobs = 5
> }
>
> Device {
> Name = “dev2"
> MediaType = “file1"
> DeviceType = "File"
> ArchiveDevice = "/mnt/folder1"
> RemovableMedia = no
> RandomAccess = yes
> AutomaticMount = yes
> LabelMedia = yes
> AlwaysOpen = no
> Autochanger = yes
> MaximumConcurrentJobs = 5
> }
>
>> On 19. Apr 2023, at 11:07, Justin Case <jus7inc...@gmail.com> wrote:
>>
>> Thanks Mat,
>> that sounds plausible, but is it even possible to have two devices for the
>> same folder on disk, which have access to the same volumes?
>> So far I believed that was impossible, but I am only a casual user of Bacula.
>>
>>> On 19. Apr 2023, at 08:00, Matlink <matl...@matlink.fr> wrote:
>>>
>>> Bacula works with disk volumes similarly as it does with tape volumes.
>>> Therefore, when truncating, it needs to mount the volume in a free drive to
>>> write some metadata on it. If you have no available drive, bacula can't
>>> operate.
>>> To solve this issue you probably need to declare a second drive like the
>>> first one.
>>> Cheers,
>>> Mat.
>>>
>>>
>>> Le 19 avril 2023 00:58:39 GMT+02:00, Justin Case <jus7inc...@gmail.com> a
>>> écrit :
>>> Hello all,
>>>
>>> while I have running backup jobs the storage is running out of space (disk
>>> file changer volumes). As the pool (pool2) that is currently written has
>>> already been truncated (manually some weeks ago), recycling of volumes does
>>> not create free space.
>>>
>>> I thus wanted to truncate volumes from a different pool (pool1) while the
>>> backup is writing to pool2:
>>> truncate pool=pool1-incr storage=storage1
>>>
>>> I get the following error (for each purged volume in pool1)(showing just 1
>>> here):
>>>
>>> Error truncating Volume “pool1-incr-vol-1109"
>>> 3908 Error reserving volume: vol_mgr.c:418 Cannot free Volume
>>> “pool2-full-vol-0703", because it is reserved by someone else.
>>>
>>> What is the reason that I cannot truncate on storage1 while a backup job is
>>> running on storage1?
>>> There is only 1 backup job running at a time.
>>>
>>> The MaximumConcurrentJobs directives are configured as follows:
>>>
>>> SD config file:
>>> -Storage1 allows 20 concurrent jobs in the SD config file.
>>> -The device allows for 5 concurrent jobs.
>>>
>>> Director config file:
>>> -The director allows 20 concurrent jobs.
>>> -Storage1 only allows 1 concurrent job in the director config file. Would
>>> this have to be configured to the same number as in the SD config file?
>>> -Is the limit of 1 here prohibiting me to truncate while the backup job is
>>> running?
>>> -The clients allow only 1 concurrent job.
>>>
>>> FD config file:
>>> -The FD allows 20 concurrent jobs.
>>>
>>> In general I want jobs to run sequentially, but truncating should be able
>>> to happen simultaneously. Is it possible to realize that?
>>> Wich MaximumConcurrentJobs directives would I need to change and how?
>>>
>>> Thanks for taking the time to consider my question,
>>> J/C
>>> 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
