Subject: Re: NetBSD2.0/sparc not ready for prime time?
To: None <port-sparc@NetBSD.org>
From: Henry Nelson <firstname.lastname@example.org>
Date: 02/10/2005 07:20:51
On Tue, Feb 08, 2005 at 04:07:35AM -0500, Chris Clymer wrote:
> I actually installed Solaris8 on one of my SparcStation5's, I believe it was
> the only remotely current versions that would even run on that box. I think
> originally it ran Solaris 2.6.1? Suffice to say, its sister box which is
> running NBSD 2.0 is quite a bit more responsive.
I run Solaris8 on a SPARCclassic. It is plenty "responsive" after turning
off all (about 8) the processes that I don't need or use that were being
started up by default. It has been running solid for a year, with pretty
heavy use during class time. I use Solaris because in a text-only environment
(CUI) the locale support for multibyte character sets is infinitely superior
The reason I'm posting, though, has more to do with the thread; not 2.0,
granted, but rather 1.6.2. I installed 1.6.2 on three IPCs, and as I posted
previously, one died on its own within a week, one's disk became so noisy
that I shut it down, and one stopped in the middle of the boot process.
I'll summarize my experiences, but the past week I've been working on the
last of these. I had very good success with running Solaris6's "format"
command off of the cd. After doing that, NetBSD installed very smoothly.
The machine has been running two days now, with zero problems.
Why did NetBSD's sysinst format the disk differently than Solaris? The
scsi probe at bootup shows the disk larger than it should be. The initial
installation before the Solaris format put a disk label with more user
available sectors than there really appear to be. I also wonder about
NetBSD reporting the total cylinders, and apparently using them, and not
leaving 2 cylinders alone as alternates as does Solaris.
As I say, more investigation. This is only a preliminary report. Still,
I very much welcome comments and/or discussion.
| day job: | http://yuba.kcn.ne.jp/biorec/nehan/henken.html