[rescue] Amiga videos online for your enjoyment

[Jonathan C. Patschke]

On Wed, 23 May 2007, Robert Brooke Gravitt wrote:

>> They DO! Admit it! I dunno what it is, but Sun systems are mega
>> slow in
>> some regards.
>
> Spoken like a man who doesn't have to watch an RS/6000 system
> ( p570 ) POST.

...which can take nearly an hour on a well-configured p570.  How nice of
IBM to make it a default to power off the system when the last LPAR is
brought down, too.  That way, even if the hardware you're moving around
is all virtual, you get to watch a full reboot...and by the time it
comes back up, you'll have forgotten to disable that little "feature".

Still, IBM BIST (POST everywhere else) is amazing stuff.  On some
hardware (the p630, the p660, and probably p5-anything), the BIST is
actually larger (both in terms of storage consumed and instructions from
end-to-end) than the AIX kernel.  No need to run offline diagnostics; if
the system comes up, it's Just Fine.

My favorite BIST story comes from a p630, into which I'd just fitted a
cryptographic accelerator.  I powered off the system[0], installed the
card, and turned the system back on.  I powered it on, and it got
partially through BIST and powered itself back off.  So, I removed the
card and brought the system back up.

When AIX had loaded, it emailed me to let me know that the previous
IPL attempt was aborted someone improperly installed a card on planar
FOO, bus BAR, slot BAZ, (location code: U1.P1.something).  It turned out
that I'd seated the card improperly.  One more attempt, and I had it up
and running.

Mmmm, I miss running IBM kit.


[0] Some combination of the system microcode level or the OS revision
     wouldn't let me hot-install that card.
-- 
Jonathan Patschke ) "If we keep our pride, though paradise is lost, we
Elgin, TX        (   will pay the price, but we cannot count the cost."
USA               )                             --Neil Peart, "Bravado"