g_vfs_done error third part--PLEASE HELP!

[Toomas Aas]

Jeremy Chadwick wrote:

> On Fri, Apr 25, 2008 at 07:59:36AM +0300, Toomas Aas wrote:
>> Willy Offermans wrote:
>>
>>> Apr 21 19:44:36 sun kernel: g_vfs_done():ar0s1g[WRITE(offset=290725074944, length=2048)]error = 5
>>> Apr 21 19:45:07 sun kernel: g_vfs_done():ar0s1g[WRITE(offset=290725074944, length=2048)]error = 5
>>> Apr 21 19:45:38 sun kernel: g_vfs_done():ar0s1g[WRITE(offset=290725074944, length=2048)]error = 5
>>> ...
>> I can only tell you that I had similar problems with FreeBSD 6.3 and ICH7R 
>> based RAID. Since I couldn't figure out how to solve them, I discarded the 
>> BIOS-based RAID and instead set up gmirror. It's been running this way for 
>> a year now and been rock solid.
> 
> Are you referring to Intel MatrixRAID?  

Yes.

> If so, there are multiple PRs
> open on problems with FreeBSD and MatrixRAID, some which have been open
> for over 2 years which include patches.  

Funny that I didn't find them when I was investigating the problem. Not 
that I'm doubting your word, just... funny.

> You wouldn't be the first
> person to ask why they haven't been committed to the tree.

Well, unfortunately I am not competent to comment on that, nor am I in 
postition to *demand* that something be committed in a volunteer project, 
since I couldn't even imagine what the consequences would be :) At least I 
found a workaround.

--
Toomas Aas