"Deathrow" down?
MG
last time I checked, DAHMER:: was not accessible, now it seems GEIN::
isn't anymore either.
- MG
Bob Koehler
MANSON is down, too, if anyone actually wanted to use it. Deathrow
is not configured as a high availability cluster, and something
common must have hit it, like power, or network.
Rob Brown
On Thu, 24 Mar 2011 at 17:47 +0100, MG wrote:
> Does anyone have an idea what happened with the "Deathrow" site?
Dunno, but I recall talk about moving the cluster sometime. There was
no recent announcement, but I suppose that could be happening now.
> The last time I checked, DAHMER:: was not accessible, now it seems
> GEIN:: isn't anymore either.
I just telnet to deathrow instead of to a specific node. That is
unreachable as well.
--
Rob Brown b r o w n a t g m c l d o t c o m
G. Michaels Consulting Ltd. (780)438-9343 (voice)
Edmonton (780)437-3367 (FAX)
http://gmcl.com/
MG
> Dunno, but I recall talk about moving the cluster sometime. There was no
> recent announcement, but I suppose that could be happening now.
As I recall it, it was planned for somewhere in June or July. Unless
Beave deviced to go ahead and do it now.
> I just telnet to deathrow instead of to a specific node. That is
> unreachable as well.
Actually, deathrow.vistech.net is GEIN::.
- MG
MG
> MANSON is down, too, if anyone actually wanted to use it.
Actually, I do. It has been down for several months now, I think I
was amongst the few who used MANSON:: and who reported it went down/
became unusable. The issues still don't seem to be known, Beave said
that he will look into it when he's going to move the cluster.
> Deathrow is not configured as a high availability cluster, and
> something common must have hit it, like power, or network.
I hope it's nothing serious either!
- MG
MG
issues (and how severe). I hear there will be a message published
within the near future about the whole situation.
- MG
Rob Brown
Maybe. Or maybe GEIN is just what you get to when you connect to
deathrow. I have never checked. I do know that at one time you did
not get GEIN when you asked for deathrow.
Hein RMS van den Heuvel
Early 2009, I donated an 2-cpu, 8-gb, 3-disk RX2600 to Deathrow
It arrived and came up fine but was never integrated as cluster member
nor independent.
:-(
This is the last I heard about it...
On Tue, Apr 7, 2009 at 3:42 PM, Da Beave
<be...@cheshire.telephreak.org> wrote:
Wooo! We got the new machine plugged in and up! Of course
it's not quite ready to go live yet (have to install a bunch of
software, etc) but we have a IP address ...
:
Hein.
MG
> Early 2009, I donated an 2-cpu, 8-gb, 3-disk RX2600 to Deathrow
> It arrived and came up fine but was never integrated as cluster member
> nor independent.
>
> :-(
>
> This is the last I heard about it...
>
> On Tue, Apr 7, 2009 at 3:42 PM, Da Beave
> <be...@cheshire.telephreak.org> wrote:
>
> Wooo! We got the new machine plugged in and up! Of course
> it's not quite ready to go live yet (have to install a bunch of
> software, etc) but we have a IP address ...
I didn't know that. I can't comment on what may have happened, but
maybe because the nodes all (used to) run V7.3-1, whereas an I64 would
--- needless to say, of course --- require V8.2 at least and gave some
problems? Maybe the fact that the systems are co-located in some data
center quite far away from him didn't give him the opportunity to set
up the rx2600 up yet, to 'connect' it even (so to speak). Just wild
guesses of course, but maybe that's also one of the reasons why he's
planning to move the entire cluster. Maybe with the current troubles,
he's going to do it a bit earlier than scheduled.
If you have any other hardware to donate in the future, you can also
consider donating to me. I'm very close to launching an I64 VMS
cluster actually. (Two nodes, at the moment, if all goes right.)
I only need to straighten out some problems, like with a replacement
System Board (UUID mismatch), install a bunch of software and so forth,
perhaps write a DCL-procedure or two as well.
- MG
Bob Koehler
>
> On Sun, 27 Mar 2011 at 13:12 +0200, MG wrote:
>
>> On 24-3-2011 18:18, Rob Brown wrote:
>>> I just telnet to deathrow instead of to a specific node. That is
>>> unreachable as well.
>>
>> Actually, deathrow.vistech.net is GEIN::.
>
> Maybe. Or maybe GEIN is just what you get to when you connect to
> deathrow. I have never checked. I do know that at one time you did
> not get GEIN when you asked for deathrow.
deathrow is a cluster alias. I don't know how dynamically it gets
managed, but it does generally map to gein. Since gein is by far
the faster machine in the cluster, and is almost never loaded down,
that makes sense even by most dynamic management algorithms.
Rob Brown
That is what I expected.
Graham Burley
switched off, it's been hanging frequently and seems to have been the
cause of the recent problems, causing GEIN to hang during startup.
MG
> Deathrow, at least GEIN& DAHMER, are back online. MANSON has been
> switched off, it's been hanging frequently and seems to have been the
> cause of the recent problems, causing GEIN to hang during startup.
It's great to hear that it's back again!
- MG
Michael Kraemer
aren't we talking about VMS clustering here?
Graham Burley
>aren't we talking about VMS clustering here?
Yes, but I expect only on Deathrow would a wedged node be left
unresolved for some months. Deathrow is not a good example of a
well-managed (or well-designed) VMS cluster, that's part of its
charm.
VAXman-
Like the charm of a rusted old '37 Chevy truck adorning the front facade of
an Appalachian homestead! :roll:
I realize the hardware has been donated for Deathrow and that it may not be
the state of the art or well maintain, but there are enough VMS savvy folks
involved there to configure it to showplace the benefits and features of a
VMScluster. Or, at least, you'd think so.
--
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG
All your spirit rack abuses, come to haunt you back by day.
All your Byzantine excuses, given time, given you away.
MG
> I realize the hardware has been donated for Deathrow and that it may not be
> the state of the art or well maintain, but there are enough VMS savvy folks
> involved there to configure it to showplace the benefits and features of a
> VMScluster. Or, at least, you'd think so.
If you have ideas, suggestions or anything else to say, perhaps you
should contact Beave?
- MG
Steve Thompson
> I realize the hardware has been donated for Deathrow and that it may not be
> the state of the art or well maintain, but there are enough VMS savvy folks
> involved there to configure it to showplace the benefits and features of a
> VMScluster. Or, at least, you'd think so.
Maybe they could even think of a less offensive name at the same time.
-s
MG
> Maybe they could even think of a less offensive name at the same time.
>
> -s
Would you think a name change is worthwhile? A lot of off-site
references might end up breaking.
Either way, if you have suggestions: Perhaps you could approach
Beave about it?
- MG
MG
was away, a SHUTDOWN was issued.)
_________
SHUTDOWN message on GEIN from user SYSTEM at _GEIN$NTY1554: 19:50:03
GEIN will shut down in 0 minutes; back up LATER. Please log off node GEIN.
SHUTDOWN
_________
I hope that later will be soon.
- MG
BR...@rabbit.hsd1.ma.comcast.net
moving from NC, USA to Jacksonville, FLA, USA.
MG
> "Beave" left a message on Deathrow NOTES, that indicates that the systems are
> moving from NC, USA to Jacksonville, FLA, USA.
Okay, I see, I missed that then. I guess he rescheduled it, since it
was originally planned for somewhere in June/July.
- MG
VAXman-
There was a message about moving the kit to the land of seizure citizens
and bugs.
VAXman-
Dunno who's da Beave is but Brad is correct. Beave left a message about
moving the systems.
Bob Koehler
> Apparently "Deathrow" was shut down today, not sure why. (While I
> was away, a SHUTDOWN was issued.)
There was a note that it was moving.
Michael Kraemer
> "Beave" left a message on Deathrow NOTES, that indicates that the systems are
> moving from NC, USA to Jacksonville, FLA, USA.
And why has it to be shutdown for that?
Aren't we talking about
"The multisite OS that just works"?
VAXman-
But at least one of the OpenVMS boxes is an emulator that is running
on the microshite OS that just never works! ;)
Richard B. Gilbert
I'm sure that the truck could accomodate a gasoline powered generator of
sufficient capacity to keep the machines up. I'm sure that if powered
up on the truck the machines would run without problems.
Maintaining network continuity would be considerably more difficult.
MG
Are you familiar with the cluster and its users? It's provided free
of charge. For "Beave" (the main proprietor and administrator) to
perform a 'live' relocation, he'd have to drive up and down at least
twice. I don't think "Beave" or anyone else involved would want to
spend that kind of effort and money in it, let alone to turn it into a
multi-site cluster. I'd say it's overkill, for the amount of users and
their typical activities.
Also, as has been said before, it's not really an HA cluster, or not in
the typical mission-critical sense obviously. Again, it's provided for
free. I can't speak for you or anyone else, but I'm very grateful for
the services that they --- entirely free of charge --- provide to me
and anyone else. It's an excellent way to become acquainted with VMS,
for those interested and potential newcomers, especially ever since the
HP Testdrive program ended.
- MG
MG
> There was a note that it was moving.
It has been brought to my attention.
- MG
Freedom on the Oceans
> > And why has it to be shutdown for that?
> > Aren't we talking about
> > "The multisite OS that just works"?
>
> I'm sure that the truck could accomodate a gasoline powered generator
> of
> sufficient capacity to keep the machines up. I'm sure that if
> powered
> up on the truck the machines would run without problems.
>
> Maintaining network continuity would be considerably more difficult.
Can I suggest using GPRS for that?
--
Tactical Nuclear Kittens
Michael Kraemer
<marc...@SPAMxs4all.nl> writes:
>
> Are you familiar with the cluster and its users? It's provided free
> of charge. For "Beave" (the main proprietor and administrator) to
> perform a 'live' relocation, he'd have to drive up and down at least
> twice. I don't think "Beave" or anyone else involved would want to
> spend that kind of effort and money in it, let alone to turn it into a
> multi-site cluster.
Nevertheless that would have been an opportunity
for a live demonstration of VMS' multisite capabilities
- if they exist at all.
Certainly better than regurgitating urban legends
of moving Police Departments.
> I'd say it's overkill, for the amount of users and
> their typical activities.
And probably HP thinks the same as far as VMS marketing is concerned.
VAXman-
>On Thu, 2011-04-28 at 10:52 -0400, Richard B. Gilbert wrote:
>> > And why has it to be shutdown for that?
>> > Aren't we talking about
>> > "The multisite OS that just works"?
>> I'm sure that the truck could accomodate a gasoline powered generator
>> sufficient capacity to keep the machines up. I'm sure that if
>> up on the truck the machines would run without problems.
>> Maintaining network continuity would be considerably more difficult.
>
>Can I suggest using GPRS for that?
I use EVDO to listen to internet radio in my moving vehicle. With the aid
of a Cradlepoint CTR500, my passengers can surf the internet by sharing my
EVDO connection. Along the US interstates there's excellent coverage too.
VAXman-
Deathrow has nothing to do with HP.
Bob Koehler
>
> And why has it to be shutdown for that?
> Aren't we talking about
> "The multisite OS that just works"?
That cluster was never configured as a high-availablility cluster,
not to mention the cost of setting up a multi-site cluster.
And I suspect from his post that Beave has limited time in which to
do the movement, not enough time to leave some systems, move others,
then repeat.
Bob Koehler
>
> Nevertheless that would have been an opportunity
> for a live demonstration of VMS' multisite capabilities
> - if they exist at all.
You mean the kind DEC, Compaq, HP, and thier customers have demonstated
several times over, already?
Or were you willing to provide the funds and manpower for the extra
equipment and work?
MG
nodes will follow the next few days, or soon.
- MG
G Cornelius
> Nevertheless that would have been an opportunity
> for a live demonstration of VMS' multisite capabilities
> - if they exist at all.
So, Michael, have you ever been anything on this group
other than a troll?
George
Michael Kraemer
<corn...@eisner.decus.org> writes:
>
> So, Michael, have you ever been anything on this group
> other than a troll?
I thought that's a prerequisite to post here.
adria...@gmail.com
li...@openmailbox.org
adrianp005--- via Info-vax <info...@rbnsn.com> wrote:
> Is Deathrow down again???
Was it up again???
VAXman-
>Is Deathrow down again???
It's been for some time now.
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG
MG
> Is Deathrow down again???
It might be permanent this time.
- MG
David Froble
Neil Rieck
IIRC, the last time this happened it was due to firewall issues at the host site which were unresolved until someone came back from vacation.
I hope this site is not permanently gone. I used it "to prove" that migration from Alpha to Itanium was going to be as simple as compile/link/run. (Everyone here already knew this but we usually forget most other members of society do not frequent this watering hole).
Neil Rieck
Kitchener / Waterloo / Cambridge,
Ontario, Canada.
http://www3.sympatico.ca/n.rieck/OpenVMS.html