Re: [Bacula-users] ERR=Cannot write block. Device is disabled.

[Kern S. <ke...@si...>]

Hello Tilman,

Hopefully your problems are now definitively solved.

Yes, I agree with you the enable should be documented.  It is most 
likely documented in the New Features chapter of the manual.  This week, 
I will verify that everything is indeed documented there and that, at 
least some if not all, of the new features are added directly to the 
manual in addition to remaining in the New Features chapter.

Best regards,
Kern


On 06/17/2018 07:45 PM, Tilman Schmidt wrote:
> Hello Kern,
>
> thanks for your answer.
> I had already successfully cleaned the drive.
> The "enable storage" command was the missing piece.
> After
>
> *enable storage=LTO-2
> 3002 Device ""LTO-2" (/dev/nst0)" enabled.
> 3004 Device ""LTO-2" (/dev/nst0)" deleted 1 alert.
>
> the drive is working fine again.
> I must admit I had missed that part in the New Features section.
>
> Btw the Bacula Console and Operators Guide for Current Version 9.0.x
> (both HTML and PDF versions) still lists only "jobs" as a possible
> keyword to the "disable" and "enable" commands.
> Entering "help enable" in bconsole reveals the "storage" keyword as well
> as three others I didn't know about.
> Perhaps the new usage could be added to the manual, and the command
> description in the help command ("Enable a job, attributes batch
> process") updated accordingly.
>
> Thanks,
> Tilman
>
> Am 17.06.2018 um 11:06 schrieb Kern Sibbald:
>> Hello,
>>
>> This is the relatively new tape Alert code in Bacula.  The messages are
>> for the most part hardware tape errors that Bacula has detected.  You
>> can look up tape alert codes, but in general, you will find that they
>> tell you the same thing that Bacula has reported.  In this case, after
>> the I/O error, Bacula checks the tape alert you configured (or you took
>> the default) and it reports tape alert 20.  This particular alert
>> generally indicates that the tape media itself is bad, but to keep from
>> contaminating the drive, Bacula has automatically disables the drive.
>> It also reported that you need to clean your drive.
>>
>> See your drive manufacturer's manual for how to clean the drive --
>> generally it is cleaned by inserting (or autoloading) a cleaning
>> cartridge.  Note, if a drive is particularly dirty it may require
>> several cleanings.  Then you can enable the drive with the "enable"
>> command in Bacula, and if I am not mistaken, all the old alert status
>> codes are cleared on the next "mount" command (even if it is already
>> mounted).  This should be documented in the Bacula new features section
>> of the document.
>>
>> By the way, after such an incident (somewhat rare for most people), I
>> would recommend to do a btape "test" command as described in the
>> manual.  Note: it needs a non-labelled tape and will write on it.
>>
>> Best regards,
>> Kern
>>
>> On 06/16/2018 11:16 PM, Tilman Schmidt wrote:
>>> My Bacula 9.0.8 installation aborted a job on an LTO-2 drive with a
>>> pretty spectacular series of error messages, claiming a defective tape
>>> and asking for a cleaning:
> [snip]
>>> As Bacula had not ejected the tape and was still blocking the drive, I
>>> had to unmount the drive in bconsole in order to do the cleaning, which
>>> was successful in clearing the cleaning indicator on the drive itself.
>>> However when I try to mount the next tape, Bacula still considers the
>>> device disabled:
>>>
>>> *mo LTO-2
>>> block.c:460 Cannot write block. Device is disabled. dev="LTO-2"
>>> (/dev/nst0)
>>> 3902 Cannot mount Volume on Storage Device ""LTO-2" (/dev/nst0)" because:
>>> Read label block failed: requested Volume "" on Tape device "LTO-2"
>>> (/dev/nst0) is not a Bacula labeled Volume, because: ERR=Cannot write
>>> block. Device is disabled. dev="LTO-2" (/dev/nst0)
>>> 3905 Device ""LTO-2" (/dev/nst0)" open but no Bacula volume is mounted.
>>> If this is not a blank tape, try unmounting and remounting the Volume.
>>>
>>> How do I clear that condition?
>>> I didn't find anything on that topic in the documentation, and Google
>>> also turned up nothing on that message.
>>>
>>> aTdHvAaNnKcSe
>>> Tilman