22 Nov
2013
22 Nov
'13
10:41 a.m.
Hi Tom,
thanks for sharing your experience with the list !
There are two 1000A Alphaservers in my collection also waiting for a revival. They had
been used by a german TV-sender to broadcast publicity. I think I'm well prepared now
when I'm going to find the time to fire the systems up regarding pitfalls known about
these machines.
Kind regards,
Pierre
?
-------------------------------------------------------------------------------
Pierre's collection of classic computers : http://classic-computing.dyndns.org/
________________________________
Von: Tom publix <ittybittybytes at gmail.com>
An: General Discussion: On-Topic and Off-Topic Posts <cctalk at classiccmp.org>
Gesendet: 23:06 Donnerstag, 21.November 2013
Betreff: Re: Alphaserver 1000 diagnosis?
I got A AS1000 on ebay as we speak. It woks fine but getting it ready I
found out a few things. TRU64 is really picky. If you have unsupported
cards or even supported cards with the wrong or outdated firmware you can
get machine checks.
Older version of BSD's Net, Free and Open will machine check. I use NetBSD
6.1 and it works fine in almost any configuration and even with cards with
outdated firmware (it will complain though)
If you have the StorageWorks backplane misconfigured it can machine check.
By the way, I had the Storage works backplane configured just to use the
qlogic 1020 controller and not the raid card, and I had concurrently 50/68
and SCA containerss and the AS1000 running Netbsd identified them all
correctly and I was able to read and write to all the disks even the SCA
one!!.
tomp
On Fri, Nov 15, 2013 at 6:52 PM, Daniel Seagraves
<dseagrav at lunar-tokyo.net>wrote:
(I think I asked this once before, but I
can't remember the result and I
can't find the thread in the archive)
I have an Alphaserver 1000 that I would like to resurrect. It periodically
hangs or reboots. When it reboots the error is usually "Machine check while
in palcode". When it hangs, the hang is so severe the halt button does not
work. Either way, it always happens within 20 minutes of booting. I ran all
of the diagnostics you get by moving the CPU card jumper (cache memory
tests, RAM tests) and they run for an hour or so with no error. (I am
assuming the diagnostic halts on error instead of just looping again, is
that correct?) During the course of these I moved the jumper to the wrong
place and wiped out SRM, so I had to reload SRM from floppy using the
failsafe loader. That worked. The SRM diagnostics will also run for as long
as I want without error. Does anyone have any idea what's wrong here or is
there some kind of XXDP-like thing I can run that will better test the
machine?