Re: SiI 3112A + Seagate HDs = still no go?

[Tejun Heo <htejun@gmail.com>]

Chris Boot wrote:
> Some interesting developments!
> 
> I installed a fresh copy of Windows, and all the VIA and nVidia and  so 
> on drivers. At some point during all this (a period of relatively  heavy 
> disk IO), the computer seemed to crash and I rebooted it. It  then 
> worked fine for a while, but during my perfmon testing it seemed  to do 
> the same thing. This time I left it for a while and it did  eventually 
> wake up again, so I'm guessing the controller is a bit  fubared. Perfmon 
> did indeed show several dips down to or very close  to 0 during the 
> write operation, with peaks up to 48 MB/sec, which is  pretty 
> respectable. So, time to replace the brand-new controller I  guess.
> 
> Now, do you think this is just my one particular controller card and  a 
> simple return would fix the problem, or is it more likely a problem  
> with the whole range? It's an Innovision EIO SATA controller: http:// 
> www.ivmm.com/eio/products/index.htm
> 
> Would it be a safer bet to go for the Adaptec controller of the same  
> variety? How reliable are they?

  I frankly don't know.  Maybe it's just one faulty controller, 
connector or whatever.  Maybe the card manufacturer screwed up 
somewhere.  I mean, the only course I took in electronics is 
introductory digital circuits which used 74xx chips and push triggered 
clock on a breadboard.  What would I know about gigahertz signaling 
error.  :-p

  Though, one thing I can say is majority of 311x controllers don't seem 
to suffer from this problem.  So, take your pick.

-- 
tejun