Netra T1-200 debugging (was: Free Ultra2 in Silicon Valley, USA)
mdh_lists at yahoo.com
Tue Nov 4 15:02:47 PST 2008
--- On Tue, 11/4/08, Marius Strobl <marius at alchemy.franken.de> wrote:
> From: Marius Strobl <marius at alchemy.franken.de>
> Subject: Re: Netra T1-200 debugging (was: Free Ultra2 in Silicon Valley, USA)
> To: "mdh" <mdh_lists at yahoo.com>
> Cc: "Mark Linimon" <linimon at lonesome.com>, freebsd-sparc64 at freebsd.org
> Date: Tuesday, November 4, 2008, 5:49 PM
> On Tue, Nov 04, 2008 at 05:09:00AM -0800, mdh wrote:
> > By the way, I'm pretty sure that there's no
> ATA controller in these. Should just be good ole SCSI. I
> mention this because I noticed Marius mentioned the ata
> driver in another message.
> Well, there in fact is an ATA controller in T1 200...
OK. I wasn't aware that those CDROMs were not SCSI-connected. Just out of curiosity, did they still show up as 't6' in /dev? ;)
> > I haven't ever worked with a T1-200 myself. It
> does indeed look like an intermediate step,
> engineering-wise, between two very well-engineered machines
> (the T1-105 and the V120).
> > Marius, I'm curious - the T1-200 has dual-eri
> interfaces on the mainboard, similar to the V100/V120?
> AFAICT the T1 200 and V120 share the exact same mainboard,
> the former is just the telco version with a different front
> bezel and not sold with an 650MHz CPU. IMO "both"
> are very
> well engineered.
> T1 105 seem more like the first try of Sun doing an 1U
> machine. They even didn't build these an own
> but recycled the CP1500 cPCI module which is just mounted
> inside the chasis together with an expansion board. IMO
> this setup and the resulting wiring makes these machines
> seem kind of fragile. From reading just this list I got
> the impression that the mezzanine RAM these machines use
> is prone to be defective.
I've never had a problem personally, but it was just that - their first try at a 1u box.
> V100 are different beasts again and seem to be designed
> with lowest-cost in mind. They have a rather small
> mainboard without a PCI-slot, not even ERI NICs but
> Davicom DM9102A and of course no SCSI. Otherwise these
> are the same generation technology as T1 200/V120, while
> T1 105 belong to the previous one.
The V100 was preceded by the "Netra X1" which was similar to the V100 but had a shorter chassis. The X1 had a V-series front bezel, but was still branded Netra. I actually liked the X1's for their size and price; they were Sun's first "really-really-cheap" SPARC-based server. It wasn't particularly reliable, sadly, but was practically disposable anyway at under $700 a pop. It came out in between the T1-105 and the V-series.
More information about the freebsd-sparc64