Re: [smartmontools-support] smartctl health check fails

[Alex S. <ml...@os...>]

On 10/10/2013 09:11 PM, Chanda Yadavalli (cyadaval) wrote:
> -----Original Message-----
> From: Alex Samorukov [mailto:ml...@os...]
> Sent: Wednesday, October 09, 2013 10:05 AM
> To: Christian Franke; Chanda Yadavalli (cyadaval)
> Cc: sma...@li...
> Subject: Re: [smartmontools-support] smartctl health check fails
>
> On 10/08/2013 11:56 PM, Christian Franke wrote:
>> This was added intentionally because issuing the SMART STATUS command
>> via this controller's SAT layer produced bogus results.
>>
>> Please remove these lines to check whether is works with recent firmware.
> I would not recommend to do this in production systems. It will possibly increment counter of "other drive errors" and at some point drive will be marked as fault. Actually SAT layer in LSI raid is a one big issue - it works very differently on different raids, but all implementations i saw are very limited. It is 100% firmware issue.
It always been randomly, but usually ~10 attempts was enough to put RAID 
in degraded state.