"amr0: bad slot x completed" and fsck_ufs hanging

Marco Beishuizen marco at beishuizen.info
Mon Nov 22 10:49:47 PST 2004


On stardate Sun, 21 Nov 2004, the wise Marco Beishuizen entered:

> On stardate Sun, 21 Nov 2004, the wise Panagiotis Christias entered:
>
>> the only hint I could find is from the amr(4) man page:
>> 
>>  amr%d: bad slot %d completed
>> 
>>  The controller reported completion of a command that the driver did not
>>  issue.  This may result in data corruption, and suggests a hardware or
>>  firmware problem with the system or controller.
>> 
>> Do you have a second controller available to test? We have the same
>> controllers in several of our servers and I would interested to find
>> out what is the problem just in case..
>
> Unfortunately it's the only controller I have so I can't test it with an 
> other controller.
>
> It looks a bit strange that it could be a hardware or firmware problem 
> because the system is brand new. But does this mean I should update my 
> firmware to a newer version?
>
> Since you don't seem to have any problems with FreeBSD on this controller I'm 
> also interested in your configuration. Did you do anything special in the 
> BIOS of the controller or something?

Looks like I've found the problem.

I reinstalled everything, including the RAID arrays on the SCSI controller. 
The first time I changed the read and write policies in the BIOS console 
and it seems that wasn't a very good idea. Now I didn't change them, and 
all seems to run fine now.

So I'm glad that the hardware is ok and I don't have to upgrade the 
firmware.

Marco

-- 
Reality is a cop-out for people who can't handle drugs.


More information about the freebsd-questions mailing list