Hello,
First off thanks for the links. I'll try those changes next week when I have a
bit more time to do the necessary testing.
I'll also try turning up the debug level to see if it does give me a bit more
information.
In the event that someone looks at this post and wonders about the TWO EOF
option. It does seem to be required with the Dell l500. The btape tests failed
when I removed it from the configuration.
Thank you Martin for taking the time to give me a hand. I will post again when
I have some results from the changes.
James
On 2011-07-07, at 12:13 PM, Martin Simmons wrote:
>>>>>> On Thu, 7 Jul 2011 10:30:36 -0600, James Woodward said:
>>
>> Hello,
>>
>> I haven't seen anything about using nsa devices to use over something like
>> an sa device. I did a bit of a search but haven't seen anything that really
>> explains that portion to me.
>
> http://www.bacula.org/5.0.x-manuals/en/main/main/Storage_Daemon_Configuratio.html#SECTION002030000000000000000
> http://www.bacula.org/5.0.x-manuals/en/problems/problems/Testing_Your_Tape_Drive.html#SECTION00413000000000000000
>
> Also, all of the examples use non-rewind devices.
>
>
>> When I set these up I did see a recommendation to use the pass through
>> devices e.g pass0, pass1, pass2, pass3. I don't remember the specifics but
>> the passthrough devices themselves seemed to be problematic.
>
> Yes, the passthrough devices shouldn't be used for data transfer. They should
> be used to control an autochanger robot.
>
>
>> I ran through all the btape tests recommended in the manual when we did the
>> initial set up and all the tests did pass.
>>
>> The Bacula config files are quite different from the older ones. I didn't
>> think this type of issue would really come down to any of these as a
>> potential problem. I will provide the config info for one of my devices from
>> the sd. They are all pretty much the same except for the Name, Drive Index,
>> and Archive Device.
>>
>> The way the message is worded it sounds like Bacula didn't think there
>> should be any information on the tape and it found some so it marked the
>> volume with error. I'm just not sure why that would happen unless it isn't
>> able to query the database properly.
>
> The message doesn't necessarily mean that. It is comparing a value in memory
> with the position on the tape, so there are other explanations. E.g. Bacula
> tried to rewind the tape but the hardware/OS refused. E.g. the hardware/OS
> returned 0 incorrectly when Bacula set the value in memory from the tape
> position in the past.
>
> It could also be a bug in Bacula.
>
> You could try running the bacula-sd with debug level 200, which might show
> more info about how the unexpected value was set.
>
>
>> If it would be helpful I could make my configs available without passwords
>> or client configs as well as the set up steps I put together when building
>> the server.
>>
>> Thanks again,
>>
>> James
>>
>> ----
>>
>> Device {
>> Name = Drive-0
>> Drive Index = 0
>> Media Type = LTO4
>> Archive Device = /dev/sa0
>> Autoselect = yes #(default)
>> Automatic Mount = yes
>> AlwaysOpen = yes #(default)
>> RemovableMedia = yes #(default)
>> RandomAccess = no
>> Requires Mount = no #(default)
>> AutoChanger = yes
>> Hardware End of Medium = no #(not default, FFSF & !HEoM is common)
>> Fast Forward Space File = yes
>> BSF at EOM = yes #(not default, necessary)
>> Backward Space Record = no #(not default)
>> Backward Space File = no #(not default)
>> TWO EOF = yes #(not default)
>> Maximum File Size = 4GB #(def is 1GB but > reduces stops to write EOF)
>> Device Type = Tape
>> Alert Command = "sh -c 'tapeinfo -f %c |grep TapeAlert|cat'"
>> Spool Directory = /spool
>> Maximum Spool Size = 3800G
>> Maximum Job Spool Size = 800G
>> }
>
> This looks OK, though you may not need the two eofs model.
>
> I use the following for an HP LTO1 drive on FreeBSD 8:
>
> Archive Device = /dev/nsa0
> AutomaticMount = yes
> AlwaysOpen = yes
> Maximum Open Wait = 10
> Offline On Unmount = no
> Hardware End of Medium = no
> BSF at EOM = no
> Backward Space Record = yes
> Backward Space File = yes
> Fast Forward Space File = yes
> TWO EOF = no
>
> Beware though that tapes written with different values of TWO EOF are not
> compatible with each other.
>
> __Martin
>
>
>>
>>
>>
>>
>> On 2011-07-07, at 9:21 AM, Martin Simmons wrote:
>>
>>>>>>>> On Thu, 7 Jul 2011 09:07:32 -0600, James Woodward said:
>>>>
>>>> Hello,
>>>>
>>>> Our older production servers are running FreeBSD 7.2 and Bacula 3.0.2.
>>>>
>>>> This particular server with the this issue is running FreeBSD 8.1 and
>>>> Bacula
>>>> 5.0.3 installed from the packages.
>>>
>>> OK, so it could be an OS bug too.
>>>
>>> Using /dev/sa2 on FreeBSD is not advisable -- can you try it with /dev/nsa2
>>> instead?
>>>
>>> What is the Device config for this drive in bacula-sd.conf? Is that
>>> identical
>>> to the production machines?
>>>
>>> Have you tried running the btape test?
>>>
>>> __Martin
>>>
>>>
>>>>
>>>> Thanks,
>>>>
>>>> James
>>>>
>>>> On 2011-07-07, at 8:41 AM, Martin Simmons wrote:
>>>>
>>>>>>>>>> On Thu, 7 Jul 2011 08:04:50 -0600, James Woodward said:
>>>>>>
>>>>>> Hello Martin,
>>>>>>
>>>>>> It's a Dell L500 and does not auto rewind. We have another 5 identical
>>>>>> ones
>>>>>> in production but they are running an older version of Bacula. This
>>>>>> seems to
>>>>>> be new to us with Bacula 5.0.3.
>>>>>
>>>>> It's true, "Expected 0, got 20" sounds like the opposite problem.
>>>>>
>>>>> Which operating system is running on the SD and is that the same in
>>>>> production?
>>>>>
>>>>> __Martin
>>>>>
>>>>>
>>>>>>
>>>>>> Thanks for the reply.
>>>>>>
>>>>>> James
>>>>>>
>>>>>>
>>>>>> On 2011-07-07, at 4:28 AM, Martin Simmons wrote:
>>>>>>
>>>>>>>>>>>> On Wed, 6 Jul 2011 08:29:46 -0600, James Woodward said:
>>>>>>>>
>>>>>>>> Hello,
>>>>>>>>
>>>>>>>> I've tried searching around a bit but I can't seem to find an answer
>>>>>>>> as to
>>>>>>>> what might cause this problem. I have qutie a few tapes that have been
>>>>>>>> marked in error due to Invalid Tape position. I've listed the Invalid
>>>>>>>> tape
>>>>>>>> position errors as well as detail about one of the affected volumes.
>>>>>>>> Has
>>>>>>>> anyone else encountered this or have any idea's as to why it might be
>>>>>>>> happening? For the life of me I can't figure out what the Expected 0
>>>>>>>> portion
>>>>>>>> refers to. I thought it might refer to volfiles but those numbers don't
>>>>>>>> always match either.
>>>>>>>> ...
>>>>>>>> Invalid Tape Position Errors
>>>>>>>> # grep -i "invalid tape position" log
>>>>>>>> 11-Mar 18:01 esqb6-sd JobId 196: Error: Invalid tape position on
>>>>>>>> volume "090137" on device "Drive-2" (/dev/sa2). Expected 0, got 20
>>>>>>
>>>>>> Is /dev/sa2 an auto-rewinding device? Bacula must use a no-rewinding
>>>>>> device
>>>>>> (normally called /dev/nsa2 on BSD).
>>>>>>
>>>>>> __Martin
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>> All of the data generated in your IT infrastructure is seriously
>>>>>> valuable.
>>>>>> Why? It contains a definitive record of application performance,
>>>>>> security
>>>>>> threats, fraudulent activity, and more. Splunk takes this data and makes
>>>>>> sense of it. IT sense. And common sense.
>>>>>> http://p.sf.net/sfu/splunk-d2d-c2
>>>>>> _______________________________________________
>>>>>> Bacula-users mailing list
>>>>>> Bacula-users AT lists.sourceforge DOT net
>>>>>> https://lists.sourceforge.net/lists/listinfo/bacula-users
>>>>>>
>>>>>>
>>>>>
>>>>> ------------------------------------------------------------------------------
>>>>> All of the data generated in your IT infrastructure is seriously valuable.
>>>>> Why? It contains a definitive record of application performance, security
>>>>> threats, fraudulent activity, and more. Splunk takes this data and makes
>>>>> sense of it. IT sense. And common sense.
>>>>> http://p.sf.net/sfu/splunk-d2d-c2
>>>>> _______________________________________________
>>>>> Bacula-users mailing list
>>>>> Bacula-users AT lists.sourceforge DOT net
>>>>> https://lists.sourceforge.net/lists/listinfo/bacula-users
>>>>>
>>>>
>>>
>>> ------------------------------------------------------------------------------
>>> All of the data generated in your IT infrastructure is seriously valuable.
>>> Why? It contains a definitive record of application performance, security
>>> threats, fraudulent activity, and more. Splunk takes this data and makes
>>> sense of it. IT sense. And common sense.
>>> http://p.sf.net/sfu/splunk-d2d-c2
>>> _______________________________________________
>>> Bacula-users mailing list
>>> Bacula-users AT lists.sourceforge DOT net
>>> https://lists.sourceforge.net/lists/listinfo/bacula-users
>>>
>>
>
> ------------------------------------------------------------------------------
> All of the data generated in your IT infrastructure is seriously valuable.
> Why? It contains a definitive record of application performance, security
> threats, fraudulent activity, and more. Splunk takes this data and makes
> sense of it. IT sense. And common sense.
> http://p.sf.net/sfu/splunk-d2d-c2
> _______________________________________________
> Bacula-users mailing list
> Bacula-users AT lists.sourceforge DOT net
> https://lists.sourceforge.net/lists/listinfo/bacula-users
>
------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2d-c2
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users
|
