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

Re: Debian on Sparc



On 17 Dec 1999, Francois Deppierraz wrote:

> Date: 17 Dec 1999 17:29:07 GMT
> From: Francois Deppierraz <francois@ctrlaltdel.ch>
> To: debian-isp@lists.debian.org
> Subject: Re: Debian on Sparc
> Newsgroups: ml.debian.isp
> Resent-Date: 17 Dec 1999 17:28:35 -0000
> Resent-From: debian-isp@lists.debian.org
> Resent-cc: recipient list not shown: ;
> 
> mc@ainsystems.com wrote:
> > Sun Netra's don't use fujitsu processors.   They use Sun Micro Electronic's CPU's in
> > the systems.  The netra xxxx models are their flag ship fault tollerant product
> > intended for NEBS environments.  I doubt the system would receive NEBS bellcore
> > certification with unreliable components.    Buying a netra with fujitsu CPU's is
> > like purchasing an intel motherboard with an Athlon on it.
> 
> I have one of the only Sun Netra which is built on a Fujitsu proc :(
> 

My 'Netra i' says:
# cat /proc/cpuinfo | grep cpu
cpu		: Fujitsu  MB86904

which runs at 110Mhz. It'a MicroSparcII compatible, AFAIK.

I've got also a SparcStation 4, running Solaris. and:
# dmesg | fgrep cpu0
cpu0: FMI,MB86904 (mid 0 impl 0x0 ver 0x4 clock 70 MHz)

It's the same CPU, at lower clock. The SparcStation 4 was bought directly
from Sun.

The 'Netra i' runs a RHL6.0 happily (their 2.2.5-22 kernel).

Comparison to modern Intel systems in unfair, this is 7 years older
technology. Used with the built-in 10base-T eth, they are good servers
for such a bandwidth (expecially a 170Mhz SS20). Based on my limited
experience, they react almost linearly to lignt and moderated loads
(2-10 clients). Modern Intel-based (low-end) systems are definitely faster
but do not show a linear degradation as the load increases. On 100Mbps
Ethernet, you should go for a Intel based server-class system.

.TM.

-- 
      ____/  ____/   /
     /      /       /			Marco Colombo
    ___/  ___  /   /		      Technical Manager
   /          /   /			 ESI s.r.l.
 _____/ _____/  _/		       Colombo@ESI.it


Reply to: