Hi,
On 5/24/2007 9:18 AM, Adam Cécile wrote:
> I'm thinking about creating an admin job that parse volumes list and
> mark all volumes not in changer as used. What would happened with bacula
> then ? Does volumes with status "used" but without any recent writing
> dates will be recycled according to their last written date ?
Yes.
But I don't suppose that would be the solution to your problem.
I'm quite sure that Bacula would not understand the two autochangers to
be the same devices, so they could not easily share volumes.
> Adam Cécile a écrit :
>> Hi,
>>
>> Not really... I made some changes on my configuration which should fix
>> the problem but it doesn't :
>>
>> ### SD conf
>> Autochanger {
>> Name = Dell-PV132T
>> Device = ULTRIUM-TD3-1
>> Device = ULTRIUM-TD3-2
>> Changer Command = "/etc/bacula/scripts/mtx-changer %c %o %S %a %d"
>> Changer Device = /dev/sg2
>> }
>> Autochanger {
>> Name = Dell-PV132T-1drive
>> Device = ULTRIUM-TD3-1
>> Changer Command = "/etc/bacula/scripts/mtx-changer %c %o %S %a %d"
>> Changer Device = /dev/sg2
>> }
>> Device {
>> Name = ULTRIUM-TD3-1
>> Drive Index = 0
>> Media Type = LTO3
>> Archive Device = /dev/nst0
>> AutomaticMount = yes;
>> AlwaysOpen = yes;
>> RemovableMedia = yes;
>> Offline On Unmount = no;
>> AutoChanger = yes
>> }
>> Device {
>> Name = ULTRIUM-TD3-2
>> Drive Index = 1
>> Media Type = LTO3
>> Archive Device = /dev/nst1
>> AutomaticMount = yes;
>> AlwaysOpen = yes;
>> RemovableMedia = yes;
>> Offline On Unmount = no;
>> AutoChanger = yes
>> }
>>
>> ### Dir conf
>> Storage {
>> Name = Autochanger
>> Address = srv-mpp-lrs
>> SDPort = 9103
>> Password = "xxx"
>> Device = Dell-PV132T
>> Media Type = LTO3
>> Autochanger = yes
>> Maximum Concurrent Jobs = 2
>> }
>> Storage {
>> Name = Autochanger-1drive
>> Address = srv-mpp-lrs
>> SDPort = 9103
>> Password = "xxx"
>> Device = Dell-PV132T-1drive
>> Media Type = LTO3
>> Autochanger = yes
>> Maximum Concurrent Jobs = 1
>> }
>>
>> As you can see I've written to autochanger ressources, but one has only
>> one drives. The same problems still occurs but now we can't tell bacula
>> haven't looked a inchanger flag because the device wasn't an autochanger...
Well, the problem is just the two autochangers... Bacula does not kow
they share the media, and so it won't load a volume marked as being in
one of them into the other.
>> Any other pointers ?
For now I can only recommend to find a setup where you don't have to
address he drives separately. I.e., create one autochanger with two
drives, use the autochanger device in the DIR only, and let Bacula
handle the details.
This will still not allow to set up jobs to use a distinct tape drive,
though.
Arno
>> May I run two different SD ? (maybe without having "always open = on")
>>
>>
>> Arno Lehmann a écrit :
>>
>>> Hi,
>>>
>>> well, I think this problem is understood now :-)
>>>
>>> On 5/23/2007 9:00 AM, Adam Cécile wrote:
>>>
>>>
>>>> Hi,
>>>>
>>>> Here is what happens :
>>>> Blogs are scheduled on the first appendable volume, without looking at
>>>> the inchanger flag. If I mark this volume as Used or Full, jobs gets
>>>> scheduled on the next appendable volume, instead of the next one
>>>> appendable and inchanger.
>>>>
>>>> #### Scheduled jobs
>>>> Scheduled Jobs:
>>>> Level Type Pri Scheduled Name Volume
>>>> ===================================================================================
>>>> Full Backup 5 23-mai-07 12:00 Backup-srv-mpp-archi-dwh
>>>> 000121
>>>> Full Backup 5 23-mai-07 12:00
>>>> Backup-srv-mpp-archi-treso 000121
>>>> Full Backup 5 23-mai-07 12:00
>>>> Backup-srv-mpp-archi-totem 000121
>>>> Full Backup 5 23-mai-07 12:00
>>>> Backup-srv-mpp-archi-filemaker 000121
>>>> Full Backup 5 23-mai-07 12:00
>>>> Backup-srv-mpp-archi-equilibre 000121
>>>> Full Backup 5 23-mai-07 20:00
>>>> Backup-srv-mpp-exchang-daily 000121
>>>>
>>>>
>>>> ### List volumes (on this Pool)
>>>> Pool: Daily
>>>> +---------+------------+-----------+---------+-------------------+----------+--------------+---------+------+-----------+-----------+---------------------+
>>>> | MediaId | VolumeName | VolStatus | Enabled | VolBytes |
>>>> VolFiles | VolRetention | Recycle | Slot | InChanger | MediaType |
>>>> LastWritten |
>>>> +---------+------------+-----------+---------+-------------------+----------+--------------+---------+------+-----------+-----------+---------------------+
>>>> | 59 | 000143 | Purged | 1 | 394,391,772,372 |
>>>> 422 | 1,209,600 | 1 | 1 | 1 | LTO3 | 2007-04-05
>>>> 08:44:03 |
>>>> | 60 | 000115 | Purged | 1 | 152,609,505,455 |
>>>> 163 | 1,209,600 | 1 | 8 | 0 | LTO3 | 2007-04-19
>>>> 13:29:21 |
>>>> | 61 | 000116 | Purged | 1 | 1,278,772,008,661 |
>>>> 1,409 | 1,209,600 | 1 | 9 | 0 | LTO3 |
>>>> 2007-04-23 13:30:52 |
>>>> | 62 | 000117 | Purged | 1 | 742,127,695,846 |
>>>> 789 | 1,209,600 | 1 | 10 | 0 | LTO3 | 2007-04-26
>>>> 21:10:42 |
>>>> | 63 | 000118 | Full | 1 | 824,915,470,096 |
>>>> 894 | 1,209,600 | 1 | 11 | 0 | LTO3 | 2007-05-09
>>>> 15:38:16 |
>>>> | 64 | 000119 | Purged | 1 | 25,754,213,310 |
>>>> 31 | 1,209,600 | 1 | 12 | 0 | LTO3 | 2007-04-27
>>>> 08:48:46 |
>>>> | 65 | 000120 | Full | 1 | 553,103,626,001 |
>>>> 577 | 1,209,600 | 1 | 13 | 0 | LTO3 | 2007-05-11
>>>> 13:59:48 |
>>>> | 66 | 000121 | Append | 1 | 1
>>>> | 0 | 1,209,600 | 1 | 14 | 0 | LTO3 |
>>>> 0000-00-00 00:00:00 |
>>>> | 91 | 000103 | Purged | 1 | 227,552,907,045 |
>>>> 250 | 1,209,600 | 1 | 10 | 0 | LTO3 | 2007-05-04
>>>> 13:58:14 |
>>>> | 92 | 000104 | Purged | 1 | 1
>>>> | 0 | 1,209,600 | 1 | 11 | 0 | LTO3 |
>>>> 0000-00-00 00:00:00 |
>>>> | 93 | 000105 | Append | 1 | 1
>>>> | 0 | 1,209,600 | 1 | 12 | 0 | LTO3 |
>>>> 0000-00-00 00:00:00 |
>>>> | 94 | 000106 | Append | 1 | 1
>>>> | 0 | 1,209,600 | 1 | 13 | 0 | LTO3 |
>>>> 0000-00-00 00:00:00 |
>>>> | 95 | 000107 | Append | 1 | 1
>>>> | 0 | 1,209,600 | 1 | 14 | 0 | LTO3 |
>>>> 0000-00-00 00:00:00 |
>>>> | 96 | 000101 | Append | 1 | 1
>>>> | 0 | 1,209,600 | 1 | 8 | 0 | LTO3 |
>>>> 0000-00-00 00:00:00 |
>>>> | 97 | 000102 | Append | 1 | 1
>>>> | 0 | 1,209,600 | 1 | 9 | 0 | LTO3 |
>>>> 0000-00-00 00:00:00 |
>>>> | 98 | 000129 | Used | 1 | 407,865,033,101 |
>>>> 427 | 1,209,600 | 1 | 8 | 0 | LTO3 | 2007-05-16
>>>> 13:26:57 |
>>>> | 99 | 000130 | Used | 1 | 210,849,249,755 |
>>>> 233 | 1,209,600 | 1 | 9 | 0 | LTO3 | 2007-05-17
>>>> 13:22:21 |
>>>> | 100 | 000131 | Used | 1 | 202,314,381,696 |
>>>> 224 | 1,209,600 | 1 | 10 | 0 | LTO3 | 2007-05-18
>>>> 13:59:23 |
>>>> | 101 | 000132 | Append | 1 | 1
>>>> | 0 | 1,209,600 | 1 | 11 | 0 | LTO3 |
>>>> 0000-00-00 00:00:00 |
>>>> | 102 | 000133 | Append | 1 | 1
>>>> | 0 | 1,209,600 | 1 | 12 | 0 | LTO3 |
>>>> 0000-00-00 00:00:00 |
>>>> | 103 | 000134 | Append | 1 | 1
>>>> | 0 | 1,209,600 | 1 | 13 | 0 | LTO3 |
>>>> 0000-00-00 00:00:00 |
>>>> | 104 | 000135 | Append | 1 | 1
>>>> | 0 | 1,209,600 | 1 | 14 | 0 | LTO3 |
>>>> 0000-00-00 00:00:00 |
>>>> | 105 | 000144 | Purged | 1 | 193,536
>>>> | 1 | 1,209,600 | 1 | 2 | 1 | LTO3 |
>>>> 2007-05-21 12:00:25 |
>>>> | 106 | 000145 | Used | 1 | 216,361,248,768 |
>>>> 253 | 1,209,600 | 1 | 3 | 1 | LTO3 | 2007-05-21
>>>> 23:30:23 |
>>>> | 107 | 000147 | Used | 1 | 219,473,114,112 |
>>>> 257 | 1,209,600 | 1 | 4 | 1 | LTO3 | 2007-05-23
>>>> 00:40:37 |
>>>> | 108 | 000146 | Append | 1 | 1
>>>> | 0 | 1,209,600 | 1 | 5 | 1 | LTO3 |
>>>> 0000-00-00 00:00:00 |
>>>> | 109 | 000148 | Append | 1 | 1
>>>> | 0 | 1,209,600 | 1 | 6 | 1 | LTO3 |
>>>> 0000-00-00 00:00:00 |
>>>> | 110 | 000149 | Append | 1 | 1
>>>> | 0 | 1,209,600 | 1 | 7 | 1 | LTO3 |
>>>> 0000-00-00 00:00:00 |
>>>> +---------+------------+-----------+---------+-------------------+----------+--------------+---------+------+-----------+-----------+---------------------+
>>>>
>>>> Which volumes are in the autochanger, which ones can be used:
>>>> See above.
>>>>
>>>>
>>> A little difficult to read with the line breaks :-)
>>>
>>> The query to "show what Bacula thinks is in changer" or whatever it's
>>> called in the distribution might be more readable.
>>>
>>>
>>>
>>>> What's the state of the SD:
>>>> Device status:
>>>> Autochanger "Dell-PV132T" with devices:
>>>> "ULTRIUM-TD3-1" (/dev/nst0)
>>>> "ULTRIUM-TD3-2" (/dev/nst1)
>>>> Device "ULTRIUM-TD3-1" (/dev/nst0) is mounted with Volume="000147"
>>>> Pool="Daily"
>>>> Slot 4 is loaded in drive 0.
>>>> Total Bytes=219,473,114,112 Blocks=3,402,050 Bytes/block=64,512
>>>> Positioned at File=257 Block=0
>>>> Device "ULTRIUM-TD3-2" (/dev/nst1) is not open.
>>>> Drive 1 status unknown.
>>>> ====
>>>> In Use Volume status:
>>>> 000147 on device "ULTRIUM-TD3-1" (/dev/nst0)
>>>>
>>>>
>>>> What happens when jobs start:
>>>> Nothing; bacula ask me to mount the volume it scheduled the jobs on
>>>>
>>>>
>>>> Could you tell me more about which informations could I give you while
>>>> running in debug mode ?
>>>>
>>>>
>>> The SD configuration and DIR storage devices would have helped here; I
>>> don't know if you already posted it, but if you did, I overlooked the
>>> relevant parts :-)
>>>
>>>
>>>
>>>> Thanks a lot for your help.
>>>>
>>>>
>>> I hope it did help you, but I'm quite sure you've got the problem now.
>>> Now to fix it :-)
>>>
>>> Actually, I think what you need is not easily possible with Bacula. See
>>> my previous post...
>>>
>>> Arno
>>>
>>>
>>>
>>>> Arno Lehmann a écrit :
>>>>
>>>>
>>>>> Hi,
>>>>>
>>>>> On 5/22/2007 8:59 AM, Adam Cécile wrote: sddddddqqsr
>>>>>
>>>>>
>>>>>
>>>>>> Hello,
>>>>>>
>>>>>> I just upgraded to 2.0.3 and I still have the problem. When no candidate
>>>>>> volume is mounted bacula do weird things.
>>>>>> Jobs are sheduled on a volume which is not in the library. If I mount a
>>>>>> good volume into the drive, jobs gets scheduled on it.
>>>>>> Any pointers ?
>>>>>>
>>>>>>
>>>>>>
>>>>> More details... your problem description is becoming less and less
>>>>> useful for me.
>>>>>
>>>>> Which volumes are in the autochanger, which ones can be used, what's the
>>>>> state of the SD, what happens when jobs start, etc. Also, consider
>>>>> running the DIR and the SD with debug output, so you can track what the
>>>>> daemons do internally, i.e. how they decide which volumes to use.
>>>>>
>>>>> Arno
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>> Arno Lehmann a écrit :
>>>>>>
>>>>>>
>>>>>>
>>>>>>> Hi,
>>>>>>>
>>>>>>> On 5/21/2007 10:33 AM, Adam Cécile wrote:
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>> Andreas Helmcke a écrit :
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>> On 21.05.2007 08:55, Adam Cécile wrote:
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>> Arno Lehmann a écrit :
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>> Hello,
>>>>>>>>>>>
>>>>>>>>>>> On 5/18/2007 9:02 AM, Adam Cécile wrote:
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>> Hi,
>>>>>>>>>>>>
>>>>>>>>>>>> Today my director was blocked.
>>>>>>>>>>>> It was asking for a volume which had inchangler flag set to no
>>>>>>>>>>>> while
>>>>>>>>>>>> there was 4 purged volume in the right pool.
>>>>>>>>>>>> I had to load the volume into the drive manually (with mtx) and
>>>>>>>>>>>> then
>>>>>>>>>>>> mount it, otherwise bacula keeps asking for the missing volume.
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>> Usually, Bacula shoould use whatever tape is available. Obviously,
>>>>>>>>>>> the
>>>>>>>>>>> catalog has to represent the actual autochanger inventory.
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>> How could this happend ? Do I missed something in my configuration
>>>>>>>>>>>> ?
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>> I don't think this is a configuration issue... most of the time, I
>>>>>>>>>>> see
>>>>>>>>>>> these problems when tapes were changed and no 'update slots'
>>>>>>>>>>> command was
>>>>>>>>>>> executed.
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>> Update slots has been run, and list media only reports the good
>>>>>>>>>> volumes
>>>>>>>>>> as "inchanger".
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>> This is a multi-drive autochanger.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>>>>> There is a query in Bacula to show the volumes that Bacula thinks
>>>>>>>>>>> are in
>>>>>>>>>>> the changer. Try this, and if it doesn't reflect the current
>>>>>>>>>>> inventory,
>>>>>>>>>>> do an 'update slots'. That should work.
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>> Do you mean list media could be right while the sql catalog may be
>>>>>>>>>> wrong ?
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>> If it doesn't, this should be a bug. You'll need to make sure
>>>>>>>>>>> you've got
>>>>>>>>>>> a recent version of Bacula running, the catalog is up to date,
>>>>>>>>>>> capture
>>>>>>>>>>> debug output for the DIR, and file a bug report. But I don't think
>>>>>>>>>>> this
>>>>>>>>>>> will be necessary :-)
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>> It really looks like being a bug... See my previous mail for more
>>>>>>>>>> informations.
>>>>>>>>>> Please notice that if I load the right tape with mtx and then mount
>>>>>>>>>> it
>>>>>>>>>> in bacula, the job gets written on this volumes.
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>> I am not sure but maybe your problems are related to the problems I
>>>>>>>>> have
>>>>>>>>> while using a multidrive autochanger (see me recent mails to this
>>>>>>>>> list).
>>>>>>>>>
>>>>>>>>> My impression is, that volume selection part of bacula sometimes gets
>>>>>>>>> confused and is choosing the wrong volume (or at least one which I
>>>>>>>>> consider wrong ;-) ).
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>> ...
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>> This looks like a serious issue. Which version are you running ? I'm on
>>>>>>>> 1.38.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>> You really should update to 2.0, or, if you dare running a beta
>>>>>>> version,
>>>>>>> even 2.1.10.
>>>>>>>
>>>>>>> In the newer versions, the device and volume selection and reservation
>>>>>>> was significantly improved.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>
>
--
IT-Service Lehmann [EMAIL PROTECTED]
Arno Lehmann http://www.its-lehmann.de
-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
