Any known issues with MPT SCSI?

Any known issues with MPT SCSI?

[Ian S. Nelson]

[-- Attachment #1: Type: text/plain, Size: 545 bytes --]

I'm running some Dell 1750s with a moderately customized 2.4.20 kernel,  
it has a couple of newer drivers but it's fairly standard.   I have 3 
identical systems that are turning up ext3 corruption fairly regularly.  
They are using the MPT/53c1030 SCSI.  2 of the 3 reported log corruption 
on a boot and mounted the root filesystem in read only.  The other is 
spitting

Dec  9 19:59:58 localhost Unexpected dirty buffer encountered at 
do_get_write_access:616 (08:02 blocknr 0)

 From time to time, I haven't rebooted it yet.

thanks,
Ian



[-- Attachment #2: Type: application/pgp-signature, Size: 252 bytes --]

Re: Any known issues with MPT SCSI?

[Stefan Kaltenbrunner]

Ian S. Nelson wrote:

> I'm running some Dell 1750s with a moderately customized 2.4.20 kernel,  
> it has a couple of newer drivers but it's fairly standard.   I have 3 
> identical systems that are turning up ext3 corruption fairly regularly.  
> They are using the MPT/53c1030 SCSI.  2 of the 3 reported log corruption 
> on a boot and mounted the root filesystem in read only.  The other is 
> spitting

FWIW: we have more than a dozen of IBM x335/345 maschines using this 
controller (using 2.4.22 and 2.4.23), and most of these server do have 
_significant_ IO-Load at times . Whe have not yet encountered anything 
near a corrupted filesystem although the controllers itself do "feel" 
quite slow.



Stefan