Hi Gurus
My original question, with the responses (Thanks you!), is below.
The general opinion is that the person I've been talking to is
mistaken, and there is no particular problem with Solaris 7 on an
Ultra 1. On the contrary, some Admins have quite high uptimes.
BTW, a couple of you asked if it was 32 or 64 bit installation -- I
did a 32 bit install.
My question:
>Dear Gurus
>
>I have an Ultra 1 that I recently got back online after an apparent
>failure of the Network card. It turned out to be a software issue
>and has been fixed, at least for now, with a fresh install of
>Solaris 7 11/99.
>
>(See "SUMMARY: No network after software install")
>
>An experienced Sun admin who I'm familiar with has told me that the
>problem is going to keep recurring. He maintains that Solaris 7 is
>incompatible with the Ultra 1, and that it'll probably run for 6
>weeks or so and then fail again. And that the only way to keep the
>box stable is to use 2.6.
>
>System:
>Sun Ultra 1 SBus (UltraSPARC 143MHz), OpenBoot 3.11, 192 MB memory installed
>Solaris 7 11/99.
>
>So my questions are:
>
>- does anyone have a solid, stable installation of Solaris 7 on an
>Ultra1? (I really don't want to use 2.6 unless I HAVE to)
>
>- is this an OpenBoot issue?
>
>At 16:50 -0500 2000/02/23, Mark Sherman wrote:
>>Robert, this sounded familiar, and i found this, it may help:
>>
>>SRDB ID: 17780
>>
>>SYNOPSIS: watch-net-all Open Boot PROM test fails with "ethernet chip
>>initialization failed" message
>>
>>DETAIL DESCRIPTION:
>>
>>Machines with OBP 3.x such as the Ultra 1, 2, 5, 10, 30, 60, 450 and Ultra
>>Enterprise servers (3000, 3500, etc) may fail Open Boot PROM tests such
>>as watch-net, watch-net-all or test net.
>
>- what do I need to do to make Solaris 7 stable on my 32-bit Ultra1?
>
>Thanks all, and I'll summarize.
Heartfelt thanks to all who responded. The responses:
At 20:28 -0500 2000/05/03, John Weekley <weekleyj@foobarlabs.org> wrote:
>Better get a new sysadmin. I've got a few Ultra 1's that exhibit none of the
>problems you're experiencing.
>There are probably thousands of Ultra 1's with 7 loaded running happily.
>Blanket statements like "Solaris 7 is incompatible with Ultra 1's" signal that
>the sysadmin needs to do a bit more research.
> > - does anyone have a solid, stable installation of Solaris 7 on an
> > Ultra1? (I really don't want to use 2.6 unless I HAVE to)
>
>Yes, until I loaded 8 on it. (Ultra 1 200E (32 bit), 256 MB Ram, Creator 3D).
>It had a max uptime of over 5 months, limited my by own idiocy.
At 08:02 +0200 2000/05/04, Erik Lembke <Erik.Lembke@force.de> wrote:
>I have to Ultra 1 both on Solaris 7, and they work (nearly) perfectly.
>One of them crashes every two or three weeks, but this also happened when
>2.6 was installed.
>(I've used the 32Bit Installation)
At 09:39 +0300 2000/05/04, Lefteris Skapetis <skapetis@intranet.gr> wrote:
> > weeks or so and then fail again. And that the only way to keep the
> > box stable is to use 2.6.
>
>Sounds silly.
>
> > - does anyone have a solid, stable installation of Solaris 7 on an
> > Ultra1? (I really don't want to use 2.6 unless I HAVE to)
>
>Ultra 1, (UltraSPARC 143MHz) OBP 3.11.1 1997/12/03 15:55 POST
>3.10.6 1996/10/18 10:19
>Solaris 7 11/99 (64-bit mode), up & running for 89 days and counting.
>
>
> >
> > - is this an OpenBoot issue?
>
>did you enable the 64-bit mode during the upgrade? i don't think
>this is an issue though.
>i don't remember seeing anything weird on the documents about the
>32/64-bit mode
>during the upgrade.
At 09:07 +0200 2000/05/04, Birger Wathne wrote:
>I have had an Ultra 1 running Solaris 7 since the OS arrived. No problems,
>except my machine had a 167MHz CPU with the bug that makes it halt-able
>from user level code when running 64-bit kernel, so I had to override in
>some config file to force it to run the 64-bit kernel.
>
>I have just upgraded it to Solaris 8 without any problems. Solaris 8
>hasn't been up for more than 2 weeks, though.
At 09:17 -0400 2000/05/04, Mark Baldwin wrote:
>I'm not sure why this person thinks that Solaris 7 is incompatible with
>Ultra 1's, but I have to disagree. Not only does Sun validate this
>configuration, I have had a Ultra 1 running Solaris 7 for over a year
>with NO problems. In fact, I have Solaris 7 on my lowly Sparc 5 for
>over a year now with no problems. I am vigilant about applying
>recommended patches, but other than that, they are two of the most
>stable boxes I have.
>
>If you want to do a test-net or watch-net on any Ultra class machine
>should first do a setenv auto-boot? false. Then reset-all. Then run
>test-net or watch-net. Otherwise you may have problems with these
>openboot commands. This is well documented by the way. Be sure to set
>auto-boot? back to true when done.
>
>Hope this helps, Mark...
At 08:24 -0500 2000/05/04, Michael Kriss wrote:
>That's ridiculous! I've had two Ultra 1's running on 2.7 for almost
>a year now
>with no problems whatsoever. I don't know exactly what version of
>OBP I have on
>these two machines but I don't suspect that to be the problem either. Is your
>Ultra 1 greater than 200 Mhz? Are you running the 64 bit kernel or the 32 bit
>kernel?
At 14:24 -0400 2000/05/04, CIsenburg@ixl.com wrote:
>You aren't by chance connected to a Cisco Catalyst?
No, I'm not.
>
>I have noticed that some Solaris NICS don't negotiate speed when
>connected... can't say that it explains your reported chip failure, but the
>failed negotiation could be reported as such...I guess...
>
>The fix, if it is the case, is:
>ndd -set /dev/hme instance 0
>ndd -set /dev/hme adv_autoneg_cap 0
>ndd -set /dev/hme adv_100fdx_cap 1
>ndd -set /dev/hme adv_100T4_cap 0
>ndd -set /dev/hme adv_100hdx_cap 0
>ndd -set /dev/hme adv_10fdx_cap 0
>ndd -set /dev/hme adv_10hdx_cap 0
>
>forces the negotiation with an infrastructure device...
Thanks again to everyone.
Best,
/Rob
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Robert Alexander
Senior Designer/Analyst/Admin
WWW Database Applications
http://www.ra1.net
"God does not subtract from one's allotted time
on Earth those hours spent flying." --Unknown
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:14:07 CDT