[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: The Last Word on UDMA on Alpha?



On Wed, 2005-01-12 at 14:34 -0800, Steve Langasek wrote:
> On Wed, Jan 12, 2005 at 11:16:24AM -0600, John Goerzen wrote:
> > I've been running my IDE disks on my 164LX ever since I got it, and
> > they've run crappy the whole time.
> 
> > I've never got the 32-bit support to work, and I'm not really sure I've
> > made DMA work either.  Turning on the 32-bit support is a sure recipe
> > for a crash, it seems.
> 
> > My 164LX has two IDE controllers: its integrated, on-botherboard one
> > (the CMD646, I believe) and a Promise 20269 PCI card.  This PCI card can
> > do UDMA133 on a PC.
> 
> > But I can't make it do that reliably on the Alpha.  I've got 2.6.9 on
> > there.
> 
> > Any ideas or suggestions?  Is IDE permenantly crapy on Alpha?  And if
> > so, why?
> 
> > Also...
> 
> > What is the final word on what journaling filesystems work on Alpha?
> > >From what I've heard here:
> 
> > ext3: OK
> > JFS: broken
> > Reiser3: broken (but it works fine for me)
> > XFS: broken
> 
> I've been using XFS on alpha for years, with no more incident than an x86.
> I've also never heard anything about it being broken on alpha except from
> you, AFAIK, but perhaps I've just been less diligent at seeking out this
> information than I should have. :)
> 
> It's also possible that the JFS problems are historical at this point; if
> someone can get more concrete information about what was broken and when
> (if) it was fixed, we can certainly still revisit the question of
> disallowing JFS in the alpha installer.

I am using JFS on a 2100A right now. under load it seems fine. If I
remember right that is when the problems were encountered.

I pummeled it too. I have LVM configured underneath it too. Seems just
fine, I am using it for a SVN repository right now.
-- 
greg, greg@gregfolkert.net

The technology that is
Stronger, better, faster:  Linux

Attachment: signature.asc
Description: This is a digitally signed message part


Reply to: