Date: Mon, 14 Feb 2005 17:47:27 -0800 (PST) From: "Jonathan A. Dama" <jd@ugcs.caltech.edu> To: Doug White <dwhite@gumbysoft.com> Cc: freebsd-amd64@freebsd.org Subject: Re: Possible problems with Broadcom BCM5704C 10/100/1000 on Tyan Motherboard Message-ID: <Pine.LNX.4.53.0502141729540.17073@groat.ugcs.caltech.edu> In-Reply-To: <20050214101303.R13953@carver.gumbysoft.com> References: <20050214161959.1FDC343D39@mx1.FreeBSD.org> <20050214101303.R13953@carver.gumbysoft.com>
next in thread | previous in thread | raw e-mail | index | archive | help
We also have these boards, I've found them unusable under
FreeBSD/5.3-STABLE with 8GB of RAM--other qualities appear to work okay.
But I even saw some infrequent problems with 6GB.
FreeBSD/amd64 is not in my opinion not actually a stable tier 1 quality
release under these configurations, too many problems remain--especially
in regards to ia32 emulation.
Exigencies of the moment forced us to forgo further debugging and adopt
linux/amd64. (Sadly, some people actually have to get work done on their
hardware...)
To anyone who wants to peg these problems on hardware, running linux these
machines have operated without fault while under a mix of high
computational and i/o load. moreover, the machines were tested
extensively using memtest+ in a controlled ambient temperature range from
60F to 80F.
This is a really lamentable situation. We've been a primarily FreeBSD
shop for 10 years now and for the past 4 years or so a pure FreeBSD shop.
Switching to linux on just these machines has been quite the headache but
I'm holding on to the hope that FreeBSD/amd64 will shape up.
FYI, most of the positive reports I've seen regarding FreeBSD and this
motherboard are 2GB setups. In my own testing that arrangement worked
_very_ well.
Addendum: The RAM timing is a bit marginal on the second processor. i.e.,
RAM that runs fine under extensive memtest+ ing has trouble
doing 400MHz DDR on the Second Processor. We ended up running
it at 333MHz DDR
-Jon
On Mon, 14 Feb 2005, Doug White wrote:
> On Mon, 14 Feb 2005, Alan Jay wrote:
>
> > I have FreeBSD 5.3 STABLE onto our new twin operteron Tyan Thunder K8S Pro
> > S2882 with 8Gb of RAM and had a reasonably stable operation for a few days we
> > installed a couple of databases one worked fine but the other kept on causing
> > the server to crash.
>
> I'm about to gain access to an S2881, which is a similar board (different
> layout but same parts).
>
> > I have searched the archive and there were issues last year but I couldn't
> > work out if these have been totally resolved?
> >
> > The adapter does work fine in low levels of loading but when pushed (it is
> > connected to a Gigabit switch) it seems to be the cause of the reboot - a what
> > appeared to be stable server with moderate Ethernet activity was fine upping
> > the activity with a new service caused regular reboots.
> >
> > There is no console message at the point of reboot to help that we have
> > spotted.
>
>
> Hm, triple fault or other hardware reset. This usually indicates bad
> hardware. Have you tried swapping the RAM between the systems and seeing
> if the problem follows? An unrecoverable ECC fault can cause a reboot,
> along with strangeness caused by temperature/power supply/etc. Or the
> board could be Just Plain Bad.
>
> Considering you have one working machine, adn this is a very popular
> board, I don't think it s abasic problem with FreeBSD and this hardware.
> The worst thing reported is interrupt routing usually.
>
> --
> Doug White | FreeBSD: The Power to Serve
> dwhite@gumbysoft.com | www.FreeBSD.org
> _______________________________________________
> freebsd-stable@freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-stable
> To unsubscribe, send any mail to "freebsd-stable-unsubscribe@freebsd.org"
>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.LNX.4.53.0502141729540.17073>