Smartd Logs

[Clovis Tristao]

Bruno Wolff III escreveu:
> On Mon, Jul 16, 2007 at 08:05:08 -0600,
>   Karl Larsen <k5di at zianet.com> wrote:
>   
>> Clovis Tristao wrote:
>>     
>>> Hi All,
>>>
>>> After installing Fedora 7, they are appearing the messages below, in
>>> log:
>>>
>>> --------------------- Smartd Begin ------------------------
>>> Currently unreadable (pending) sectors detected:
>>>    /dev/sdb - 48 Time(s)
>>>    2987 unreadable sectors detected
>>>
>>> Offline uncorrectable sectors detected:
>>>    /dev/sda - 48 Time(s)
>>>    10 offline uncorrectable sectors detected
>>>    /dev/sdb - 48 Time(s)
>>>    2987 offline uncorrectable sectors detected
>>>
>>> ---------------------- Smartd End -------------------------
>>> Why appear this messages? I'm worried.
>>>
>>>
>>> Another question, because the system changed the recognition of the
>>> HD-IDE of hda for sda? Thanks a lot,
>>>
>>> Clóvis
>>>
>>>
>>>
>>>       
>>    First, with Fedora 7 /dev/hdb1 becomes /dev/sdb1. You are looking at 
>> the root messages and mine have the same sort of thing saying it can't 
>> find something. I expect it is a bug but I would have trouble writing 
>> one for this.
>>
>>    Do not worry if your system comes up and works :-)
>>     
>
> That is really bad advice.
>
> Clóvis, needs to find out what is causing the problem and if he can't, he
> should at the very least be prepared to have to replace his disk on short
> notice using a relatively current back up of his system.
>
> 3000 sectors is only a small fraction of a modern disk so it is not hard to
> believe that the system could still appear to be functioning with that
> many bad sectors. However, if the message is correct, his disk is essentially
> toast and he should immediately do a back up and replace the disk.
>
> I have seen some reports on the smartmontools list where a recent kernel
> change may be causing some problems, but I don't think the symptoms
> correspond to what Clóvis is seeing.
>
> Clóvis, my adice to you is to immediately back up your system (or at least
> the files you care about). After that, what you do next depends on how valuable
> your data is. Probably you should just buy another disk and set the old
> one aside in case any of your important files were corrupted and you need
> to make heroic recovery attempts later. If you have more time than money and
> your data isn't very important, you might try running badblocks off a live
> cd and using it to rewrite and test all of the sectors on your disk, and
> then use smartctl to look at whether the sectors were reallocated and
> whether or not you disk thinks that too many were to continue using the disk.
>
>   
Dear Bruno,

Already I am in my as record, the previous one gave to problem of bad 
sector.

The current record is new.

Models:
New Disk: ata1.00: ATA-7: SAMSUNG SP0411N, TW100-11, max UDMA/100 
ata1.00: 78242976 sectors, multi 16: LBA48 ata1.01: ata_hpa_resize 1: 
sectors = 78165360, hpa_sectors = 67108869

Old Disk: ata1.01: ATA-6: ST340014A, 3,06, max UDMA/100 ata1.01: 
78165360 sectors, multi 16: LBA48

Maybe bug in Smartd? I'm googling for any informatio for smartd lot of 
message.

Thanks for your reply,

Clóvis 

-- 
Clovis Tristao - UNICAMP/Faculdade de Engenharia Agricola
Administrador de Redes - Secao de Informatica (SINFO)
E-mail: clovis at agr.unicamp.br http://www.feagri.unicamp.br
Fone(0xx19) 35211031-35211038 ou FAX(55xx19) 35211005/35211010