READ_DMA, WRITE_DMA errors

[Steve]

Paul Mather wrote:

>One common thread in my case is that
>all ran some kind of software RAID (gvinum or gmirror), though not all
>of my software RAIDed machines exhibited the DMA problems leading me to
>think perhaps it was a hardware/load/disk combination problem.  
>  
>
I do not use RAID at all, so, not common for me.

>Anyway, as well as 5-STABLE, I also run a 6-CURRENT system that suffered
>the problem.  Happily, after the ATA Mk.III merge, the situation
>improved a LOT.  I occasionally still get the error reported, but it is
>not fatal, unlike before (where the drive would be detached, breaking my
>geom_mirror, necessitating a lengthy background rebuild).  
>  
>
Well, that's good news, I just hope that is a widespread fix, there 
seems to be different issues, and, hopefully, the rewrite intentionally 
or unintentionally resolves them all! Sounds like in your case, it's 
almost 100%. An occasional error (we get watchdog timeouts on network) 
is not bad as long as it doesn't destroy the FS, obviously, we want 
zero, but, things happen. It's quite conceivable that 1 error per day IS 
a hardware issue. But, in our case, with 4 machines and the corruption, 
not the case!
 
Steve