SS10 died on power outage-- wont boot
Marc J. Fraioli
We have an SS10 Model 30 here, which, alas, is not on an UPS. There
was a power outage due to a thunderstorm here last night, and the
system bought it. When I got in this morning, it was on, but the
screen was black, and it wouldn't answer ping. So I turned it off,
waited a bit, and turned it back on. It came up with all sorts of
errors I had never seen before, and wouldn't boot. One thing I
noticed is that it said:
Ethernet address: fe:fe:fe:fe:fe:fe, Host ID: fefefefe
Then lower down,
The IDPROM contents are invalid
It would then do some of its normal boot stuff, and eventually get to:
Boot: vmunix
Size: 1515520+460808+215056 bytes
Data Access Exception
ok
I first tried using L1-N as I turned it on, and this gave the desired
"resetting NVRAM" message, but it still failed to boot. So, I
tried booting from the Solaris 1.1.1B CD (which is what was installed
on the machine), and got the following:
The IDPROM contents are invalid
... (normal stuff deleted) ...
boot: vmunix
size: 860160+2318280+65032 bytes
Invalid format type (254) in NVRAM
machine type 0xffffffff in NVRAM
panic: No known machine type configured in!
Data Access Exception
ok
Has anyone seen this sort of thing before? Is it possible to fix
it from the PROM monitor, or does it need physical repairs? Any
info much appreciated.
Marc Fraioli
m...@clark.net
Himanshu Gohel
and a 4/470 console in addition to 630MP.
Last week we had multiple power outages (about 3 in a span of a few
minutes). I noticed that the two Sparc 10s were UNAFFECTED by the
power outage, while the Sparc 2s and both the servers went down.
The Sparc 10s continued happily, just waiting for our home server
to come back up...
Sun has done something right with the Sparc 10s!
(BTW, one Sparc 10 was sharing the same surge strip as
the 4/470, and I believe the other Sparc 10 was sharing it
with one of the Sparc 2s)
--
Himanshu Gohel, Go...@csee.usf.edu
U of South Florida, Tampa, FL 33620-3140
Lupe Christoph
> We have an SS10 Model 30 here, which, alas, is not on an UPS. There
>was a power outage due to a thunderstorm here last night, and the
>system bought it. When I got in this morning, it was on, but the
>screen was black, and it wouldn't answer ping. So I turned it off,
>waited a bit, and turned it back on. It came up with all sorts of
>errors I had never seen before, and wouldn't boot. One thing I
>noticed is that it said:
> Ethernet address: fe:fe:fe:fe:fe:fe, Host ID: fefefefe
>Then lower down,
> The IDPROM contents are invalid
Actually, the NVRAM has been wiped out.
> ...
>I first tried using L1-N as I turned it on, and this gave the desired
>"resetting NVRAM" message, but it still failed to boot. So, I
>tried booting from the Solaris 1.1.1B CD (which is what was installed
>on the machine), and got the following:
Doesn't help. This just resets part of the NVRAM, like the environment.
> boot: vmunix
> size: 860160+2318280+65032 bytes
> Invalid format type (254) in NVRAM
> machine type 0xffffffff in NVRAM
> panic: No known machine type configured in!
The machine type is derived from the hostid, and 0xfe is not
a valid machine type.
>Has anyone seen this sort of thing before? Is it possible to fix
>it from the PROM monitor, or does it need physical repairs? Any
>info much appreciated.
You can program the NVRAM with a programmer that accepts NVRAMs.
But to do this, you would have to have a good NVRAM to copy, plus,
of course the programmer.
I'd try to talk to Sun Service and find out how much they charge
for a new NVRAM. I'm assuming you don't have a maintenance
contract, coz then you would probably have called them already.
--
| lu...@alanya.isar.muc.de (private, German IN) | Disclaimer: |
| lu...@cic.de (company, XLink) | This is an unofficial |
| suninfo!alanya!lupe (for Sun Germany) | opinion of Christoph & |
| Non sunt multiplicanda entia praeter necessitatem. | Imschweiler Consulting |
robert...@gmail.com
commands in for setting variables and you can use them to set the nvram. I
have had to do this in the distant past.