🚀 go-pugleaf

On Thu, Nov 15, 2012 at 03:00:18PM +0100, Ignatios Souvatzis wrote:
> Is this list moderated?

The list may be, but I took my last system (AlphaPC164SX) out of service
four years ago.  A sad day.  It lives in a cardboard box now.  Until the
next time I move.  Or anyone would like it.

-- 
 ___ ___  .   .  ___
  \    /  |\  |\ \
  _\_ /__ |-\ |-\ \__

On 11/15/2012 09:00 AM, Ignatios Souvatzis wrote:
> Is this list moderated?

  Alive?  Certainly.  Moderated?  Who knows, but I doubt it.

             -Dave

-- 
Dave McGuire, AK4HZ
New Kensington, PA

On 11/15/2012 03:29 PM, Izaac wrote:
> On Thu, Nov 15, 2012 at 03:00:18PM +0100, Ignatios Souvatzis wrote:
>> Is this list moderated?
> The list may be, but I took my last system (AlphaPC164SX) out of service
> four years ago.  A sad day.  It lives in a cardboard box now.  Until the
> next time I move.  Or anyone would like it.
>
Have one here that is still in use:

% uname
NetBSD speedy.ludd.ltu.se 4.0_BETA2 NetBSD 4.0_BETA2 (SPEEDY) #3: Sat 
Jan 30 09:31:42 MET 2027 
lps@speedy.ludd.ltu.se:/usr/src/sys/arch/alpha/compile/SPEEDY alpha
% uptime
  4:32PM  up 337 days,  8:33, 5 users, load averages: 0.16, 0.10, 0.09
% /sbin/sysctl hw.model
hw.model = Digital AlphaPC 164LX 533 MHz

-- Ragge

> Is this list moderated?

Moderated?  As far as I know (which is a view from the outside), no
more than any of the other (public) NetBSD lists.

Alive?  Not very. :(

					Mouse

Hi Dave.

Yep. Still running on a DS10 here.

Regards,
Al.

On Thu, 15 Nov 2012, Dave McGuire wrote:

> On 11/15/2012 09:00 AM, Ignatios Souvatzis wrote:
>> Is this list moderated?
>
>  Alive?  Certainly.  Moderated?  Who knows, but I doubt it.
>
>             -Dave
>
>

-- 
  --
  Al Boyanich
  adb -w -P "world> " -k /dev/meta/galaxy/ksyms /dev/god/brain

   I suppose this makes me a bit of a traitor, but I re-purposed my 
DS10L not long ago; I installed VMS on it and it's now a HECnet node.

   I'm still running NetBSD/alpha on a 3000 model 300 though.

            -Dave

On 11/15/2012 03:36 PM, ..I'd rather be coding ASM! wrote:
>
> Hi Dave.
>
> Yep. Still running on a DS10 here.
>
> Regards,
> Al.
>
> On Thu, 15 Nov 2012, Dave McGuire wrote:
>
>> On 11/15/2012 09:00 AM, Ignatios Souvatzis wrote:
>>> Is this list moderated?
>>
>>  Alive?  Certainly.  Moderated?  Who knows, but I doubt it.
>>
>>             -Dave
>>
>>
>


-- 
Dave McGuire, AK4HZ
New Kensington, PA

I'm still actively using NetBSD/alpha on a AlphaPC 164, mainly as a file
server and place to run misc *nix stuff on (primary computer runs
Windows most of the time).

NetBSD yerfable.azeotrope.org 6.99.7 NetBSD 6.99.7 (YERFABLE) #15: Mon 
May 21 22:53:35 CDT 2012 
root@vmbsd.azeotrope.org:/usr/obj.alpha/sys/arch/alpha/compile/YERFABLE 
alpha

On Thu, Nov 15, 2012 at 04:41:58PM +0100, Anders Magnusson wrote:
> On 11/15/2012 03:29 PM, Izaac wrote:
> >On Thu, Nov 15, 2012 at 03:00:18PM +0100, Ignatios Souvatzis wrote:
> >>Is this list moderated?
> >The list may be, but I took my last system (AlphaPC164SX) out of service
> >four years ago.  A sad day.  It lives in a cardboard box now.  Until the
> >next time I move.  Or anyone would like it.
> >
> Have one here that is still in use:

I do have one too, mostly to track -current ...

njoly@thanos [~]> uname -a
NetBSD thanos.sis.pasteur.fr 6.99.15 NetBSD 6.99.15 (GENERIC-$Revision: 1.346 $) #15: Mon Nov 19 18:21:44 CET 2012  njoly@thanos.sis.pasteur.fr:/local/src/NetBSD/obj.alpha/sys/arch/alpha/compile/THANOS alpha
njoly@thanos [~]> /sbin/sysctl hw.model
hw.model = Digital Personal WorkStation 500au

And run the regression tests :

  http://www.netbsd.org/~njoly/alpha-atf/

-- 
Nicolas Joly

Biology IT Center
Institut Pasteur, Paris.

slow on the uptake, but still here.

alpha-wise, I'm down to the 1000A, a 3000/400, and some multias of
various repair.  The 1000A continues to be my primary machine; the rest
are collecting dust.

hw.model = AlphaServer 1000A 5/400
 8:52AM  up 623 days, 15:57, 15 users, load averages: 0.19, 0.17, 0.11
NetBSD arwen.poofy.goof.com 4.99.25 NetBSD 4.99.25 (ARWEN) #2: Sun Jul 22 17:51:27 PDT 2007  agrier@arwen.poofy.goof.com:/var/obj/ARWEN alpha

Updating it to NetBSD-6 (or at least to 5) has been on the todo list for
a while, but there's some outstanding NFS-related patches from Michael
Hitch that don't appear to have been merged yet.
(http://gnats.netbsd.org/cgi-bin/query-pr-single.pl?number=35448)

The 1000A 5/400 is still usable for my purposes.  it's certainly more
stable and faster than a pentium pro contemporary.  updating the aging
storage subsystem away from SCSI (add gig-E for iSCSI or add a SATA
controller for local disk) in an attempt to cut power consumption seems
like it would be beating a dead horse.

part of me realizes I should just suck it up and migrate all the
services over to (potentially virtual) amd64 systems, but there are
still aspects of the DEC systems (like BIOS vs SRM, chassis
construction, motherboard layout) that I feel are still a level above
commodity x86 boxes.

I keep the 3000/400 with the intention of testing, but with only 128MB,
it's quite ram starved for modern workloads.  it also eats over 220W
while idle, so I don't keep it on much.

the multias are pathetic, not stable, but they don't take up much space,
and I have higher priority (larger) systems to dispose of first...  :)

I would think a more modern alpha (EV68CB or later) with plenty of
memory and modern disk would be good for at least another half decade,
but they seem to be pretty rare.

any advice (or merely consolation) is welcomed.  :)

-- 
  Aaron J. Grier | "Not your ordinary poofy goof." | agrier@poofygoof.com

On 12/16/2012 07:11 PM, Aaron J. Grier wrote:
> alpha-wise, I'm down to the 1000A, a 3000/400, and some multias of
> various repair.  The 1000A continues to be my primary machine; the rest
> are collecting dust.

  I'm amazed that the 1000A is still running.  They're the only Alphas
I've ever found to be less reliable than the Multia.  I'm glad to hear
at least ONE of them was decent.

> The 1000A 5/400 is still usable for my purposes.  it's certainly more
> stable and faster than a pentium pro contemporary.  updating the aging
> storage subsystem away from SCSI (add gig-E for iSCSI or add a SATA
> controller for local disk) in an attempt to cut power consumption seems
> like it would be beating a dead horse.

  They're capable of quite a bit of work, despite industry's effort to
"obsolete" them.  ("obsolete" has been verbed!)

  But...why move away from SCSI?  I'll bet even older SCSI drives will
last a good bit longer than the consumer-grade drek flooding the market
now.  And non-consumer-grade SATA/SAS drives cost a fortune.  It'll take
a good long while to recoup that kind of cost in reduced power consumption.

> part of me realizes I should just suck it up and migrate all the
> services over to (potentially virtual) amd64 systems, but there are
> still aspects of the DEC systems (like BIOS vs SRM, chassis
> construction, motherboard layout) that I feel are still a level above
> commodity x86 boxes.

  Ugh, why?  I parse "commode" in "commodity".  That hardware is
crap...unless you buy really high-end stuff, but you can buy a new car
for what some of those machines cost.  And in the end you still have to
deal with crap like the BIOS.

  My next major infrastructure upgrade will be from UltraSPARC-III+ to
UltraSPARC-T2.  Going 2-3 years between reboots, I'm not in too much of
a rush to even do that.

> the multias are pathetic, not stable, but they don't take up much space,
> and I have higher priority (larger) systems to dispose of first...  :)

  Eh.  Multias.  I had such high hopes for those machines when I first
read about them.

> I would think a more modern alpha (EV68CB or later) with plenty of
> memory and modern disk would be good for at least another half decade,
> but they seem to be pretty rare.

  They're showing up on the market occasionally.  It's only a matter of
time.

  Man I'd bet that even current NetBSD, which has a good bit of bloat,
would run like a scalded dog on one of those machines.  Zooom!

            -Dave

-- 
Dave McGuire, AK4HZ
New Kensington, PA

On Sun, 16 Dec 2012 20:01:49 -0500
Dave McGuire <mcguire@neurotica.com> wrote:

>   But...why move away from SCSI?  I'll bet even older SCSI drives will
> last a good bit longer than the consumer-grade drek flooding the market
> now.  And non-consumer-grade SATA/SAS drives cost a fortune.  It'll take
> a good long while to recoup that kind of cost in reduced power consumption.

Agreed. I've only looked into modern stuff to get more space. I still
have several IBM DCHS04's (*) running that have been powered on 24/7
for at least eight years now, with decent load on them too. They just
won't die. And even modern NetBSD still fits in 4 gigabytes w/ swap.


As for modern 'drek', I can vouch for the Western Digital RE3. SATA,
yes, but they perform well and don't cost an arm and a leg. Needless to
say, not quite as battle proven as the DCHS but so far, so good -
several years of load on a medium-busy fileserver without a failure.



*-http://www.ebay.com/sch/i.html?_trksid=p5197.m570.l1313&_nkw=ibm+dchs04&_sacat=0&_from=R40


-- 
Aaron B. <aaron@zadzmo.org>

On Dec 17, 2012, at 8:14 AM, Aaron B. <aaron@zadzmo.org> wrote:

> On Sun, 16 Dec 2012 20:01:49 -0500
> Dave McGuire <mcguire@neurotica.com> wrote:

You guys must live somewhere where electricity is really cheap.

>>  But...why move away from SCSI?  I'll bet even older SCSI drives will
>> last a good bit longer than the consumer-grade drek flooding the market
>> now.  And non-consumer-grade SATA/SAS drives cost a fortune.  It'll take
>> a good long while to recoup that kind of cost in reduced power consumption.
> 
> Agreed. I've only looked into modern stuff to get more space. I still
> have several IBM DCHS04's (*) running that have been powered on 24/7
> for at least eight years now, with decent load on them too. They just
> won't die. And even modern NetBSD still fits in 4 gigabytes w/ swap.
> 
> 
> As for modern 'drek', I can vouch for the Western Digital RE3. SATA,
> yes, but they perform well and don't cost an arm and a leg. Needless to
> say, not quite as battle proven as the DCHS but so far, so good -
> several years of load on a medium-busy fileserver without a failure.
> 
> 
> 
> *-http://www.ebay.com/sch/i.html?_trksid=p5197.m570.l1313&_nkw=ibm+dchs04&_sacat=0&_from=R40
> 
> 
> -- 
> Aaron B. <aaron@zadzmo.org>

-- thorpej

   Not really, but this is how I make my money, so I don't mind spending
some of it to run reliable stuff.  Hardware failures are one thing I've
never had to worry about.

  That said, my new place, near Pittsburgh, has significantly better
power prices (~$0.095/kWh) than my old place in Florida, at ~$0.13/kWh.

              -Dave

On 12/17/2012 11:58 AM, Jason Thorpe wrote:
> You guys must live somewhere where electricity is really cheap.
> 
>>>  But...why move away from SCSI?  I'll bet even older SCSI drives will
>>> last a good bit longer than the consumer-grade drek flooding the market
>>> now.  And non-consumer-grade SATA/SAS drives cost a fortune.  It'll take
>>> a good long while to recoup that kind of cost in reduced power consumption.
>>
>> Agreed. I've only looked into modern stuff to get more space. I still
>> have several IBM DCHS04's (*) running that have been powered on 24/7
>> for at least eight years now, with decent load on them too. They just
>> won't die. And even modern NetBSD still fits in 4 gigabytes w/ swap.
>>
>>
>> As for modern 'drek', I can vouch for the Western Digital RE3. SATA,
>> yes, but they perform well and don't cost an arm and a leg. Needless to
>> say, not quite as battle proven as the DCHS but so far, so good -
>> several years of load on a medium-busy fileserver without a failure.
>>
>>
>>
>> *-http://www.ebay.com/sch/i.html?_trksid=p5197.m570.l1313&_nkw=ibm+dchs04&_sacat=0&_from=R40
>>
>>
>> -- 
>> Aaron B. <aaron@zadzmo.org>
> 
> -- thorpej
> 


-- 
Dave McGuire, AK4HZ
New Kensington, PA

On Mon, 17 Dec 2012 08:58:12 -0800
Jason Thorpe <thorpej@shagadelic.org> wrote:

> 
> You guys must live somewhere where electricity is really cheap.
>

Yes, but, it's still expensive enough. For that very reason, I've been
looking for a low power not x86 based platform that's stable running
NetBSD. 

The Dreamplug comes close to my specifications - I wish it had more
ram, but really I hate the form factor. Whatever I use to replace my
old RISC boxen needs to be rack mounted.

-- 
Aaron B. <aaron@zadzmo.org>

On Sun, Dec 16, 2012 at 08:01:49PM -0500, Dave McGuire wrote:
> I'm amazed that the 1000A is still running.  They're the only Alphas
> I've ever found to be less reliable than the Multia.  I'm glad to hear
> at least ONE of them was decent.

I've never met anybody else with a 1000A -- any suggestions on things to
look out for, common failures, etc?

>   But...why move away from SCSI?

roughly ordered: capacity, performance, availability, physical space,
and power consumption / waste heat.

capacities of non-SAS SCSI drives seem to be capped at 300GB, and it's
unclear to me if anybody is still in production of non-SAS SCSI drives.
sure, I can stockpile 36GB drives, but a few spare 2TB drives is a lot
smaller stack for the same storage space...

theoretically it should be possible to bridge between "classic" SCSI and
SAS (and by superset, SATA,) but I don't know where to find such
devices.  clues welcome.

> I'll bet even older SCSI drives will last a good bit longer than the
> consumer-grade drek flooding the market now.

certainly a lot of the old drives keep going!  I have an RZ56 and RZ57
in my DECstation that just seem to keep going.  unfortunately for these
drives, longevity isn't the only metric I'm concerned about.

> And non-consumer-grade SATA/SAS drives cost a fortune.

so do high capacity SCSI and SCA drives.  :)

> It'll take a good long while to recoup that kind of cost in reduced
> power consumption.

pulling out my kill-a-watt, shuffling some power cords, and exercising
redundant power supplies, I see that the external disk tray on my 1000A
is pulling a bit over 120W while idle.  over the course of a year, that
costs me roughly $125 to keep running at $0.12/kWh.  that certainly
covers the cost of a gig-E card, and I have 80GB of RAIDed free space on
another machine that I could export via iSCSI... (now could I get it all
configured for $80 of my time investment.  hee hee.)

> Ugh, why [suck it up and migrate all the services to amd64]?  I parse
> "commode" in "commodity".  [amd64] hardware is crap...

where else to go?

the same concerns as storage apply: capacity, performance, availability,
physical space, and power consumption / waste heat.

I don't recall how long it took the last time I built the NetBSD world
on my 1000A, but on a five-year-old x86 system at work, it takes about
20 minutes.

for kicks, I shuffled some disk, added some swap, and kicked off a
niced native build of NetBSD-6 on my 1000A.  hopefully it'll take less
than a day.  :)

> unless you buy really high-end stuff, but you can buy a new car for
> what some of those machines cost.

a lot of DEC gear cost as much as a new car when it was new, too.  a
better question is can I find anything (from any architecture) that has
comparable reliability.

> And in the end you still have to deal with crap like the BIOS.

EFI is indeed a horror show, but is it bad enough to forgo the other
benefits?  for me the answer is no, so long as I don't forget how it
does suck, and remember to give sh*t to the co-workers responsible when
the opportunities arise...  :)

>   My next major infrastructure upgrade will be from UltraSPARC-III+ to
> UltraSPARC-T2.  Going 2-3 years between reboots, I'm not in too much of
> a rush to even do that.

that's definitely how it should be, and something the computing industry
seems to have forgotten.  flaky hardware?  virtualize it across multiple
systems.  flaky software?  design your application for redundancy.

damn kids don't know what a lawn is anymore.  :)

-- 
  Aaron J. Grier | "Not your ordinary poofy goof." | agrier@poofygoof.com

>> Ugh, why [suck it up and migrate all the services to amd64]?  I
>> parse "commode" in "commodity".  [amd64] hardware is crap...
> where else to go?

Depends on what you're after.

If you just want computrons, then yeah, you want amd64...or possibly
ARM or some such if low power is an important consideration.  And
there's probably no point in your being on this list.

I'm here because I like the Alpha.  I have Alphas, which I sometimes
run (and would like to run more), rather than amd64s.  I have/run them
because I want to.  I enjoy it in a way I don't enjoy running on peecee
hardware.  Much the same is true of my SPARCs.  Dreamcast.  Shark.  Etc.

There is a small pragmatic benefit, in that running on a variety of
machines impels me to write more portable code.  It's only small,
though, both because I've mostly internalized the attitudes that lead
to portable code anyway and because "everybody" is running linux on
x86_64 and there's thus "no point" in being portable to anything else.
(Yes, as the quotes imply, I don't actually agree with those "facts",
nor the points of view they lead to.)  More to the point, the pragmatic
benefit can mostly be obtained with only some three or four machines,
not the list I have ({32,64}-bit, {big,little}-endian, alignment
{in,}sensitive, that covers most of it).

> the same concerns as storage apply: capacity, performance,
> availability, physical space, and power consumption / waste heat.

Yeah.  If those are more important to you than whatever considerations
got you running alphas, then I'd suggest offering the alphas here, or
on classiccmp or some such, and running amd64s instead.  You'll
probably be happier - and whoever ends up with the alphas almost
certainly will be.

But if you have other considerations, reasons you want to run alphas
instead of peecees, then they are your answer to the above, and they
can guide you in deciding whether it's better, for you, to look for a
way to use "modern" drives on your alphas or to stockpile "36G" or
"72G" drives or the like, or what.

/~\ The ASCII				  Mouse
\ / Ribbon Campaign
 X  Against HTML		mouse@rodents-montreal.org
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B

On Mon, 17 Dec 2012, Aaron J. Grier wrote:

> On Sun, Dec 16, 2012 at 08:01:49PM -0500, Dave McGuire wrote:
>> I'm amazed that the 1000A is still running.  They're the only Alphas
>> I've ever found to be less reliable than the Multia.  I'm glad to hear
>> at least ONE of them was decent.
>
> I've never met anybody else with a 1000A -- any suggestions on things to
> look out for, common failures, etc?

   I have one.  I haven't had it on for some time (too busy with too many 
other things).

   One major problem I ran into is that the 1000A has two PCI busses, and 
doesn't do the swizzle thing with the interrupts.  Plugging a PCI card 
into the 2nd bus slots doesn't work.  I was working on some changes to the 
1000A support so cards would work in the 2nd bus slots, but haven't done 
anything with them for a while.  I don't know if I'd have any time to do 
anything until after I retire (which isn't that far away).

Mike

--
Michael L. Hitch			mhitch@montana.edu
Computer Consultant
Information Technology Center
Montana State University	Bozeman, MT	USA

On Dec 18, 2012, at 2:57 AM, Aaron J. Grier <agrier@poofygoof.com> wrote:

> On Sun, Dec 16, 2012 at 08:01:49PM -0500, Dave McGuire wrote:
>> I'm amazed that the 1000A is still running.  They're the only Alphas
>> I've ever found to be less reliable than the Multia.  I'm glad to hear
>> at least ONE of them was decent.
> 
> I've never met anybody else with a 1000A -- any suggestions on things to
> look out for, common failures, etc?


Sad to say, but I recently got finished wiping the hard drives of a 1000A to get it prepped for surplus. :-(

The system ran like a champ 24/7 literally for many many years whilst its PC brethren would struggle and flake out.  Unfortunately, its demise was signalled by one of its SCSI hard drives failing, which was all the justification that was needed for the Powers That Be that didn't like "that big hulking monstrosity" to send it on its way. :-(

In a continuing nod to diversity, though, I'll be replacing it with a PPC-based Xserve G5. :-)

Cheers,

Paul.

On Dec 18, 2012, at 10:44 AM, "Michael L. Hitch" <mhitch@lightning.msu.montana.edu> wrote:

> On Mon, 17 Dec 2012, Aaron J. Grier wrote:
> 
>> On Sun, Dec 16, 2012 at 08:01:49PM -0500, Dave McGuire wrote:
>>> I'm amazed that the 1000A is still running.  They're the only Alphas
>>> I've ever found to be less reliable than the Multia.  I'm glad to hear
>>> at least ONE of them was decent.
>> 
>> I've never met anybody else with a 1000A -- any suggestions on things to
>> look out for, common failures, etc?
> 
>  I have one.  I haven't had it on for some time (too busy with too many other things).
> 
>  One major problem I ran into is that the 1000A has two PCI busses, and doesn't do the swizzle thing with the interrupts.  Plugging a PCI card into the 2nd bus slots doesn't work.  I was working on some changes to the 1000A support so cards would work in the 2nd bus slots, but haven't done anything with them for a while.  I don't know if I'd have any time to do anything until after I retire (which isn't that far away).


Another problem I discovered during my time using a 1000A with NetBSD is that it lost the ability to reboot via "shutdown -r" circa NetBSD 5: http://mail-index.netbsd.org/port-alpha/2009/06/19/msg000314.html.  It made remote updates somewhat less fun than before.  Not sure if that got fixed...

Cheers,

Paul.

On Sun, 16 Dec 2012, Aaron J. Grier wrote:

> Subject: Re: Is this list alive?

   It is still alive, but not too many people left with Alphas I would 
guess.

> alpha-wise, I'm down to the 1000A, a 3000/400, and some multias of
> various repair.  The 1000A continues to be my primary machine; the rest
> are collecting dust.

   I'm still running a PC164 at home, and one of my CS20 systems is running 
in my office.  I have a 1000A, but haven't powered it up recently. 
Rennovations to the building, including an "updated and improved" HVAC 
system have rather decreased the cooling in my office, so I don't have as 
many machines running as I used too, plus I don't seem to have much spare 
time to play with the alphas.  I also no longer have access to the ES40 
and ES45 systems we used to have at work (and never even got to try much 
of anything with the ES47 systems).  The ES40 was nice because it had 4 
cpus and could show up multi-processor problems faster than my CS20, and 
it had the remote console capability.

Mike


--
Michael L. Hitch			mhitch@montana.edu
Computer Consultant
Information Technology Center
Montana State University	Bozeman, MT	USA

On 12/18/2012 09:16 AM, Paul Mather wrote:
> Sad to say, but I recently got finished wiping the hard drives of a
> 1000A to get it prepped for surplus. :-(

    Take it home!

> The system ran like a champ 24/7 literally for many many years whilst
> its PC brethren would struggle and flake out.  Unfortunately, its
> demise was signalled by one of its SCSI hard drives failing, which
> was all the justification that was needed for the Powers That Be that
> didn't like "that big hulking monstrosity" to send it on its way.
> :-(

  ROFL!!  The 1000A is one of the smaller Alphas.  "These kids today, I
just don't know.."

> In a continuing nod to diversity, though, I'll be replacing it with a
> PPC-based Xserve G5. :-)

  Nice. :-)

            -Dave

-- 
Dave McGuire, AK4HZ
New Kensington, PA

On 12/18/2012 02:57 AM, Aaron J. Grier wrote:
>> I'm amazed that the 1000A is still running.  They're the only Alphas
>> I've ever found to be less reliable than the Multia.  I'm glad to hear
>> at least ONE of them was decent.
> 
> I've never met anybody else with a 1000A -- any suggestions on things to
> look out for, common failures, etc?

  I've had six 1000 and 1000A machines.  On five, the power supplies
died.  On the sixth, the CPU board went south.

> theoretically it should be possible to bridge between "classic" SCSI and
> SAS (and by superset, SATA,) but I don't know where to find such
> devices.  clues welcome.

  That would be interesting indeed.  I've done similar things with
FibreChannel.  It's amazing where the SCSI protocol shows up. :)

>> I'll bet even older SCSI drives will last a good bit longer than the
>> consumer-grade drek flooding the market now.
> 
> certainly a lot of the old drives keep going!  I have an RZ56 and RZ57
> in my DECstation that just seem to keep going.  unfortunately for these
> drives, longevity isn't the only metric I'm concerned about.

  Yeah. :-(

> pulling out my kill-a-watt, shuffling some power cords, and exercising
> redundant power supplies, I see that the external disk tray on my 1000A
> is pulling a bit over 120W while idle.  over the course of a year, that
> costs me roughly $125 to keep running at $0.12/kWh.  that certainly
> covers the cost of a gig-E card, and I have 80GB of RAIDed free space on
> another machine that I could export via iSCSI... (now could I get it all
> configured for $80 of my time investment.  hee hee.)

  I've considered putting together a honking big Solaris box (well,
another one) and setting it up as a ZFS-based NAS serving up iSCSI.  I
know lots of people who have done that with great success.  You can use
crappy drives but ZFS handles the crappiness.

>> Ugh, why [suck it up and migrate all the services to amd64]?  I parse
>> "commode" in "commodity".  [amd64] hardware is crap...
> 
> where else to go?

  UltraSPARC for big, ARM for small.

>>   My next major infrastructure upgrade will be from UltraSPARC-III+ to
>> UltraSPARC-T2.  Going 2-3 years between reboots, I'm not in too much of
>> a rush to even do that.
> 
> that's definitely how it should be, and something the computing industry
> seems to have forgotten.  flaky hardware?  virtualize it across multiple
> systems.  flaky software?  design your application for redundancy.
> 
> damn kids don't know what a lawn is anymore.  :)

  Yup. :-(  All I know is there will be at least ONE corner of the
computing world that is sane, at least until I go to the big datacenter
in the sky.

               -Dave

-- 
Dave McGuire, AK4HZ
New Kensington, PA

Mine (an AlphaStation 5/266) died from a power surge. I'd love to get another, but even eBay seems mostly dried up.

I can't say I would use it for much, but I love having other architectures around for compile testing. If my C ran on x86, amd64, sparc32, sparc64, and alpha, it would run on anything (I just added a DNARD to the collection... In 2012)


--
Sean Davis / IT Security Engineer
Sent from my phone


On Dec 18, 2012, at 10:01, "Michael L. Hitch" <mhitch@lightning.msu.montana.edu> wrote:

> On Sun, 16 Dec 2012, Aaron J. Grier wrote:
> 
>> Subject: Re: Is this list alive?
> 
>  It is still alive, but not too many people left with Alphas I would guess.
> 
>> alpha-wise, I'm down to the 1000A, a 3000/400, and some multias of
>> various repair.  The 1000A continues to be my primary machine; the rest
>> are collecting dust.
> 
>  I'm still running a PC164 at home, and one of my CS20 systems is running in my office.  I have a 1000A, but haven't powered it up recently. Rennovations to the building, including an "updated and improved" HVAC system have rather decreased the cooling in my office, so I don't have as many machines running as I used too, plus I don't seem to have much spare time to play with the alphas.  I also no longer have access to the ES40 and ES45 systems we used to have at work (and never even got to try much of anything with the ES47 systems).  The ES40 was nice because it had 4 cpus and could show up multi-processor problems faster than my CS20, and it had the remote console capability.
> 
> Mike
> 
> 
> --
> Michael L. Hitch            mhitch@montana.edu
> Computer Consultant
> Information Technology Center
> Montana State University    Bozeman, MT    USA

On 12/18/2012 01:11 PM, Sean Davis wrote:
> Mine (an AlphaStation 5/266) died from a power surge. I'd love to get
> another, but even eBay seems mostly dried up.

  5/266...What model, specifically?  And where are you located?  I have
some Alpha hardware here that can possibly be re-homed.

> I can't say I would use it for much, but I love having other
> architectures around for compile testing. If my C ran on x86, amd64,
> sparc32, sparc64, and alpha, it would run on anything (I just added a
> DNARD to the collection... In 2012)

  Portability is a good thing, especially now with ARM poised to take
over pretty much everything.

               -Dave

-- 
Dave McGuire, AK4HZ
New Kensington, PA

On Sun, 2012-12-16 at 16:11 -0800, Aaron J. Grier wrote:
> the multias are pathetic, not stable....

Perhaps... but they're *cute* :-)

I have one (well, more than one, but the others are for parts).  I've
played around with NetBSD 4 and 5 on it, and I'm getting ready to try
again with 6.0.

I had noticed some apparent stability problems -- random compile errors
and segfaults while building a kernel, never reproducible.  I was going
to try swapping some parts around, figuring it was some marginal
hardware somewhere.

But if it's a general Multia issue, I don't want to waste time on it.

So -- is anyone else with a Multia finding it's "not stable"?

Pathetic is another matter... that kernel build runs overnight.  But I
just want a small box that doesn't take much power to use as a firewall,
and that doesn't take a monster machine.

Upping the cooling helps.  I cut the thermistor on mine, and it worked
fine until the 2.5" SCSI drive died, and I haven't replaced it yet.

-Dustin

On Thu, Dec 20, 2012 at 11:48 PM, Bill Roman <netbsd@songdog.eskimo.com> wrote:
> On Sun, 2012-12-16 at 16:11 -0800, Aaron J. Grier wrote:
>> the multias are pathetic, not stable....
>
> Perhaps... but they're *cute* :-)
>
> I have one (well, more than one, but the others are for parts).  I've
> played around with NetBSD 4 and 5 on it, and I'm getting ready to try
> again with 6.0.
>
> I had noticed some apparent stability problems -- random compile errors
> and segfaults while building a kernel, never reproducible.  I was going
> to try swapping some parts around, figuring it was some marginal
> hardware somewhere.
>
> But if it's a general Multia issue, I don't want to waste time on it.
>
> So -- is anyone else with a Multia finding it's "not stable"?
>
> Pathetic is another matter... that kernel build runs overnight.  But I
> just want a small box that doesn't take much power to use as a firewall,
> and that doesn't take a monster machine.

On Thu, 20 Dec 2012 21:48:42 -0800
Bill Roman <netbsd@songdog.eskimo.com> wrote:

> I had noticed some apparent stability problems -- random compile errors
> and segfaults while building a kernel, never reproducible.
Usualy thats a symptome of failing RAM. Possibly as trivial as a bad
contact on one of the SIMMs / DIMMs. Sometimes this can be solved by
cleaning contacts and reseating everything..
-- 


\end{Jochen}

\ref{http://www.unixag-kl.fh-kl.de/~jkunz/}

On 12/21/2012 12:48 AM, Bill Roman wrote:
>> the multias are pathetic, not stable....
> 
> Perhaps... but they're *cute* :-)
> 
> I have one (well, more than one, but the others are for parts).  I've
> played around with NetBSD 4 and 5 on it, and I'm getting ready to try
> again with 6.0.
> 
> I had noticed some apparent stability problems -- random compile errors
> and segfaults while building a kernel, never reproducible.  I was going
> to try swapping some parts around, figuring it was some marginal
> hardware somewhere.
> 
> But if it's a general Multia issue, I don't want to waste time on it.
> 
> So -- is anyone else with a Multia finding it's "not stable"?

  I did, when I ran a few.  I have one now that I haven't fired up in a
long time.  But Back In The Day(tm), before I realized it was a crap
design, I had lots of overheating problems with them, and later the
infamous "fried 74F623" issue.  I replaced about half a dozen of those.

  If you haven't put a heat sink on your Multia's 74F623 chip, I would
suggest doing so when you can.

              -Dave

-- 
Dave McGuire, AK4HZ
New Kensington, PA

seems were getting alive after some hibernation :)

multias were thought as "thin"-clients (yeah, with alpha-cpu....)

they also have often troubles with the memories. this might explain
compile and / or segfaults. those machines are no servers!

74F623 might also be an issue, never forget for what those machines were
built...



On Fri, 2012-12-21 at 14:04 -0500, Dave McGuire wrote:
> On 12/21/2012 12:48 AM, Bill Roman wrote:
> >> the multias are pathetic, not stable....
> > 
> > Perhaps... but they're *cute* :-)
> > 
> > I have one (well, more than one, but the others are for parts).  I've
> > played around with NetBSD 4 and 5 on it, and I'm getting ready to try
> > again with 6.0.
> > 
> > I had noticed some apparent stability problems -- random compile errors
> > and segfaults while building a kernel, never reproducible.  I was going
> > to try swapping some parts around, figuring it was some marginal
> > hardware somewhere.
> > 
> > But if it's a general Multia issue, I don't want to waste time on it.
> > 
> > So -- is anyone else with a Multia finding it's "not stable"?
> 
>   I did, when I ran a few.  I have one now that I haven't fired up in a
> long time.  But Back In The Day(tm), before I realized it was a crap
> design, I had lots of overheating problems with them, and later the
> infamous "fried 74F623" issue.  I replaced about half a dozen of those.
> 
>   If you haven't put a heat sink on your Multia's 74F623 chip, I would
> suggest doing so when you can.
> 
>               -Dave
> 

On Fri, Dec 21, 2012 at 12:48 AM, Bill Roman <netbsd@songdog.eskimo.com> wrote:
> On Sun, 2012-12-16 at 16:11 -0800, Aaron J. Grier wrote:
>> the multias are pathetic, not stable....
>
> Perhaps... but they're *cute* :-)
>
> I have one (well, more than one, but the others are for parts).  I've
> played around with NetBSD 4 and 5 on it, and I'm getting ready to try
> again with 6.0.
>
> I had noticed some apparent stability problems -- random compile errors
> and segfaults while building a kernel, never reproducible.  I was going
> to try swapping some parts around, figuring it was some marginal
> hardware somewhere.
>
> But if it's a general Multia issue, I don't want to waste time on it.
>
> So -- is anyone else with a Multia finding it's "not stable"?
>
> Pathetic is another matter... that kernel build runs overnight.  But I
> just want a small box that doesn't take much power to use as a firewall,
> and that doesn't take a monster machine.

Bill,

I have a Multia, and they definitely are stable, if you manage them
properly.  The main issue with them, as is well documented in the
NetBSD Multia FAQ, is heat.  First, the fans suck, or rather don't
suck enough.  See
http://www.netbsd.org/ports/alpha/multiafaq.html#overheat.  Second,
there is an IC on the back of the board that runs rather hot, and can
cook itself.  See
http://www.netbsd.org/ports/alpha/multiafaq.html#heat-death.

I'm rather fond of these machines, and they are far from the slowest
machines supported by -current.  If you want to see slow, try a
DECstation 5000/20 (which I also have!).

- Alex

On Tue, Dec 18, 2012 at 04:53:52AM -0500, Mouse wrote:
> If you just want computrons, then yeah, you want amd64...or possibly
> ARM or some such if low power is an important consideration.  And
> there's probably no point in your being on this list.

I received some first- and second-generation opteron sunfire systems
from a friend who was retiring them.  computronically, they are great,
and the 4-way x2200m2 system builds the amd64 NetBSD-6 world in less
than an hour-and-a-quarter, but the service processor + BIOS combination
is a frankenstein horror.  undering the new layer of paint and server
facade, it's still a PC.

sunfires sure as hell aren't alphas.  maybe the HP superdome stuff is at
the same level as DEC was, but I have my doubts.  it seems to me like
the current generation of "high-end" x86 systems were architected by
people who came up through the x86 ranks and never dealt with even
workstation grade SPARC, MIPS, VAX, or Alpha.

I am watching ARM's growth into server space with interest.  I
previously had a colocated 30W atom system, and the price/performance
ratio was very good, but it was all consumer-grade commodity parts.  the
power supply failed on me after roughly a year, and the motherboard
doesn't support ECC.  I'm interested in what ARM can bring to the table.
FWIW, my sparc 2s were around 40W...

> I'm here because I like the Alpha.  I have Alphas, which I sometimes
> run (and would like to run more), rather than amd64s.  I have/run them
> because I want to.  I enjoy it in a way I don't enjoy running on
> peecee hardware.  Much the same is true of my SPARCs.  Dreamcast.
> Shark.  Etc.

the more I learn about and deal with Intel/AMD hardware, the sadder I am
about Alpha's demise.  I see some of the ideas percolating through
Intel, but dealing with the "top-of-the-line" and still seeing missing
functionality that was implemented in decades-old hardware is a
disappointment.  (it also means there potential opportunity...)

> > the same concerns as storage apply: capacity, performance,
> > availability, physical space, and power consumption / waste heat.
> 
> Yeah.  If those are more important to you than whatever considerations
> got you running alphas, then I'd suggest offering the alphas here, or
> on classiccmp or some such, and running amd64s instead.  You'll
> probably be happier - and whoever ends up with the alphas almost
> certainly will be.

I did make a call out when I got rid of the last load-o-DEC.  there were
no takers for the systems; just some people interested in parts.  (that
reminds me I need to box some things up...)

> But if you have other considerations, reasons you want to run alphas
> instead of peecees, then they are your answer to the above, and they
> can guide you in deciding whether it's better, for you, to look for a
> way to use "modern" drives on your alphas or to stockpile "36G" or
> "72G" drives or the like, or what.

while I haven't had any abjectly horrible experiences with large SATA
drives yet, I have secured a stockpile of SCA drives and a new disk
array, so I can at least keep existing in-use systems going for a while
longer.

-- 
  Aaron J. Grier | "Not your ordinary poofy goof." | agrier@poofygoof.com

> I did make a call out when I got rid of the last load-o-DEC.  there
> were no takers for the systems; just some people interested in parts.

I can believe it.  I've passed up machines I'd love to have for many
and various reasons, from lack of physical space to shipping cost. :(

/~\ The ASCII				  Mouse
\ / Ribbon Campaign
 X  Against HTML		mouse@rodents-montreal.org
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B