Dave McGuire <mcguire at neurotica.com> writes:
On 12/25/2012 03:05 PM, Cory Smelosky wrote: >> Brian (Hechinger) and I
once discussed the possibility of implementing >> a similar clustering
system within NetBSD. I think just implementing >> the SCS protocol
itself would be a good start. I wonder if there's >> enough
documentation out there to do it. > > What'd be really interesting is
joining that to a VMS cluster. ;)
It would be a huge amount of work, and I'm not sure it'd be practical.
There are a lot of things in VMS that just don't have a mappable
counterpart in the UNIX world. RMS, for instance.
I've had my head deep in the bowels of RMS for the past 4 years. It's
not something that I believe would ever see the light of day on a non-
VMS O/S notwithstanding its appearance in the ontogeny of VMS.
--
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG
Well I speak to machines with the voice of humanity.
Found the problem, STUPI isn't listing you as a known circuit:
{GORVAX$} mcr ncp tell stupi show know circ
Known Circuit Volatile Summary as of 25-DEC-2012 22:53:42
Circuit State Loopback Adjacent
Name Routing Node
QNA-0 on 59.60 (R29GWA)
QNA-0 59.10 (SOL)
QNA-1 on
TCP-0-0 on 2.1 (LEGATO)
TCP-0-1 on -starting
TCP-0-2 on 19.41 (SG1)
TCP-0-3 on -starting
So no area 61...
Sampsa
On 26 Dec 2012, at 00:31, Dave McGuire <mcguire at neurotica.com> wrote:
On 12/25/2012 05:22 PM, sampsa at mac.com wrote:
Ah, the problem is that both area 59 and 21 (Peter and Brian) aren't up - so we can't crawl them.
Well they don't seem to be up from the point of view of GORVAX and MIM...
I can see MIM just fine from here. My appears to be going through
Peter (area 59).
My tunnel to Brian appears to be up; I can ping 52.1, but none of his
other machines appear to be reachable from here right now.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
Well, I think we should leave the format alone and just add LAT= and LONG= to the location field.
It's not hard to parse :)
Sampsa
On 26 Dec 2012, at 00:50, Rob Jarratt <robert.jarratt at ntlworld.com> wrote:
I think LAT/LONG is unambiguous and also gives you the opportunity to be
more precise if you want to. We can keep "location" as a more informal
description though.
Regards
Rob
-----Original Message-----
From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE]
On Behalf Of Bob Armstrong
Sent: 25 December 2012 22:25
To: hecnet at Update.UU.SE
Subject: RE: [HECnet] HECnet mapping project
I'd just add a LAT/LONG field before the notes one
FWIW, the existing INFO.TXT format already has a "location" field - it's
City
and Country, but you can just geolocate the city center.
Bob
Cory Smelosky <b4 at gewt.net> writes:
On 25 Dec 2012, at 15:06, Dave McGuire <mcguire at neurotica.com> wrote:
On 12/25/2012 03:02 PM, Cory Smelosky wrote: >> Ah. Is any sort of
encryption done for cluster-over-IP? >=20 > Oh heck no. That sort of
overhead's resultant latency would destroy > the performance of a
cluster on all but the very fastest of machines.
Ah, right, I'm forgetting that clustering sends a lot of data.
Not necessarily. However, depending on what you are doing, DLM (Distributed
Lock Management) traffic and T/MSCP (Tape/Mass Storage Control Protocol) can
become heavy consumers of your ethernet cluster connect.
--
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG
Well I speak to machines with the voice of humanity.
Dave McGuire <mcguire at neurotica.com> writes:
On 12/25/2012 02:56 PM, Cory Smelosky wrote: >>> Will it be a problem
that all of these systems are DECnet-OSI? >> >> Clustering does not
depend on DECnet or TCP/IP! > > Really? I thought clustering over
ethernet used DECnet.
Nope...SCS, which is Systems Communications Services if memory serves.
(is that right, Brian?) Completely unrelated to DECnet.
Dunno which "Brian" you refer but SCS it is.
--
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG
Well I speak to machines with the voice of humanity.
I think LAT/LONG is unambiguous and also gives you the opportunity to be
more precise if you want to. We can keep "location" as a more informal
description though.
Regards
Rob
-----Original Message-----
From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE]
On Behalf Of Bob Armstrong
Sent: 25 December 2012 22:25
To: hecnet at Update.UU.SE
Subject: RE: [HECnet] HECnet mapping project
I'd just add a LAT/LONG field before the notes one
FWIW, the existing INFO.TXT format already has a "location" field - it's
City
and Country, but you can just geolocate the city center.
Bob
Cory Smelosky <b4 at gewt.net> writes:
On 25 Dec 2012, at 15:00, sampsa at mac.com wrote:
=20 > On 25 Dec 2012, at 21:56, Cory Smelosky <b4 at gewt.net> wrote: >=20
=20 >>> Clustering does not depend on DECnet or TCP/IP! >>=20 >>
Really? I thought clustering over ethernet used DECnet. >>=20 >=20 >
Nope, it's own LAN protocol.
Ah. That would explain why it's still functioning despite DECnet being =
hilariously broken on the simulated alpha. ;)
How's it broken?
--
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG
Well I speak to machines with the voice of humanity.
You're right.
Was the link up last night?
Sampsa
On 26 Dec 2012, at 00:29, Bob Armstrong <bob at jfcl.com> wrote:
Ah, the problem is that both area 59 and 21 (Peter and Brian) aren't up ...
Well they don't seem to be up from the point of view of GORVAX and MIM...
LEGATO is talking to Peter and STUPI (59.58).
... don't know about area 21, though.
Bob
On 26 Dec 2012, at 00:31, Dave McGuire <mcguire at neurotica.com> wrote:
On 12/25/2012 05:22 PM, sampsa at mac.com wrote:
Ah, the problem is that both area 59 and 21 (Peter and Brian) aren't up - so we can't crawl them.
Well they don't seem to be up from the point of view of GORVAX and MIM...
I can see MIM just fine from here. My appears to be going through
Peter (area 59).
My tunnel to Brian appears to be up; I can ping 52.1, but none of his
other machines appear to be reachable from here right now.
The crawler uses NCP SHOW KNOW CIRC, and this is what it looks like from GORVAX and MIM:
BTW, Brian wrote the crawler, so he's in a better position to answer questions.
I just graph the output :)
---- SNIP ----
{GORVAX$} mcr ncp show know circ
Known Circuit Volatile Summary as of 25-DEC-2012 22:38:29
Circuit State Loopback Adjacent
Name Routing Node
QNA-0 on 12.2 (BENDER)
QNA-0 44.21 (NIKKEL)
QNA-0 19.40 (SGC)
QNA-0 2.1 (LEGATO)
QNA-0 11.2 (MAISA)
QNA-0 47.556 (KUHAVX)
QNA-0 19.41 (SG1)
QNA-0 6.1 (STAR69)
QNA-0 1.13 (MIM)
QNA-0 5.1023 (A5RTR)
TCP-0-0 on -starting
TCP-0-2 on 2.1 (LEGATO)
TCP-0-3 on 3.171 (NUK1C)
TCP-0-6 on -starting
TCP-0-7 on 7.60 (BITXOV)
TCP-1-9 on 19.41 (SG1)
TCP-4-7 on 47.556 (KUHAVX)
TCP-5-9 on -starting
{GORVAX$} mcr ncp tell mim show know circ
Known Circuit Volatile Summary as of 25-DEC-2012 22:38:50
Circuit State Loopback Adjacent
Name Routing Node
UNA-0 on 44.21 (NIKKEL)
UNA-0 12.2
UNA-0 5.1023 (A5RTR)
UNA-0 6.1 (STAR69)
UNA-0 19.40 (SGC)
UNA-0 8.400 (GORVAX)
UNA-0 2.1 (LEGATO)
UNA-0 19.41 (SG1)
UNA-0 47.556 (KUHAVX)
UNA-0 11.2 (MAISA)
UNA-0 1.350 (ERSATZ)
UNA-0 1.250 (TARDIS)
On 12/25/2012 05:22 PM, sampsa at mac.com wrote:
Ah, the problem is that both area 59 and 21 (Peter and Brian) aren't up - so we can't crawl them.
Well they don't seem to be up from the point of view of GORVAX and MIM...
I can see MIM just fine from here. My appears to be going through
Peter (area 59).
My tunnel to Brian appears to be up; I can ping 52.1, but none of his
other machines appear to be reachable from here right now.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
