On 2011-12-31 02.59, Robert Armstrong wrote:
The one "problem" with it, is that machines that aren't online also
can't be queried for that file, so it might be a bit limited from that
point of view.
That's why the file format allows it to contain information about multiple machines. You'd list all your machines, powered up or not, in the single INFO.TXT on your gateway machine.
Cool. I might have forgotten some details. Looking at MIM::INFO.TXT, I realize that it does mention the machine name, but I don't know for sure if the file on MIM:: follows whatever format INFO.TXT should be in nowadays.
We should try to get a formal definition for that file in writing somewhere.
Johnny
Exactly! What Bob said... :-)
-Steve
-----Original Message-----
From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On
Behalf Of Robert Armstrong
Sent: Friday, December 30, 2011 9:00 PM
To: hecnet at Update.UU.SE
Subject: Re: [HECnet] Re: Topology 28-Dec-2011
The one "problem" with it, is that machines that aren't online also
can't be queried for that file, so it might be a bit limited from that
point of view.
That's why the file format allows it to contain information about
multiple machines. You'd list all your machines, powered up or not, in
the single INFO.TXT on your gateway machine.
Bob
The one "problem" with it, is that machines that aren't online also
can't be queried for that file, so it might be a bit limited from that
point of view.
That's why the file format allows it to contain information about multiple machines. You'd list all your machines, powered up or not, in the single INFO.TXT on your gateway machine.
Bob
The file can handle multiple machines. Just look at SLAVE::INFO.TXT or
LEGATO::INFO.TXT or even SGC::INFO.TXT. Each of these files contains an
entire DECnet area of machines.
-Steve
-----Original Message-----
From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On
Behalf Of Johnny Billquist
Sent: Friday, December 30, 2011 8:55 PM
To: hecnet at Update.UU.SE
Subject: Re: [HECnet] Re: Topology 28-Dec-2011
On 2011-12-31 02.48, Steve Davidson wrote:
Johnny,
You are talking about the INFO.TXT file that is located in the default
DECnet directory.
Right. I even have one on MIM::
I just couldn't remember the name or format of the file, but figured
someone else probably would. Little did I realized that others had
already responded with the same information. :-)
The one "problem" with it, is that machines that aren't online also
can't be queried for that file, so it might be a bit limited from that
point of view.
Johnny
-Steve
-----Original Message-----
From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On
Behalf Of Johnny Billquist
Sent: Friday, December 30, 2011 8:46 PM
To: hecnet at Update.UU.SE
Subject: Re: [HECnet] Re: Topology 28-Dec-2011
Someone (I don't remember who) had a suggestion a year or two ago,
that
machines should have a file in the default DECnet directory, with some
basic information about the node in a defined format, and the file
with
a defined name, so that anyone could easily extract and figure out
information about atleast running nodes.
Maybe that idea could be expanded upon?
Seems very tedious for people to constantly mail updates... Although I
could of course take additional information as I'm already keeping a
nodename database in Datatrieve. No big deal to add additional
information if we'd like that... Then I can generate any kind of
reports
based on this. As can others (if they can get remote Datatrieve to
work
against RSX.)
Johnny
On 2011-12-30 15.59, hvlems at zonnet.nl wrote:
Rok, the spreadsheet that I distributed also has a sheet called
'hosts'.
I'm willing to expand that list though I have the basty feeling it's
going to take a lot more time than the project itself.
The information is not readily available thru NCP. Possibly NCL has a
relevant command but if so my ncl is *very* rusty and it seems I'm
pretty much the only phase V owner.
NCP gives limited information: the set exec node and sho exec (char)
commands only tell you the os name. For vms it tells you what
platform.
The excel file contains this information for each node in a single
cell:
<address>-<name>-<hardware>-<os name>
Mark the hyphens: that way I can manipulate the csv file easily on
vms
with f$element ;-)
So anyone who mails me node data in the format listed gets data
included in the file. What do you think?
Hans
-----Original Message-----
From: Rok Vidmar<rok.vidmar at gmail.com>
Sender: owner-hecnet at Update.UU.SE
Date: Fri, 30 Dec 2011 15:40:56
To:<hecnet at update.uu.se>
Reply-To: hecnet at Update.UU.SESubject: Re: [HECnet] Re: Topology
28-Dec-2011
Sorry, I'm confused. Is it your plan to keep a manually updated
list of
all of HECnet? Are you looking for data on all known nodes, even
end
nodes
and ones that don't normally run 24x7? And how much data do you
want
for
each node? OS? Hardware type? Geographic location?
That's fine with me and I'm not complaining in the least if
that's
the
case! OTOH, I thought you were making a list of the active routing
nodes,
which is an entirely different matter.
Well, both. I generated list of active nodes and I believe it
will
not be too
hard to manually update it from time to time.
I can add whatever interesting information and I can add any
number of
editors. Any volunteers? Hans?
--
Regards, Rok
On Dec 30, 2011, at 5:37, Johnny Billquist <bqt at softjar.se> wrote:
The last time I was joined to HECnet was 11 years ago!!
Long time no see. :-)
It's good to be back!
Gha! I was trying to remember who might have been around back then. So that was you... I think that makes you the first place outside my own experimentation that was connected to HECnet. :-)
I believe that is correct. I don't remember anyone else being connected then.
Serial ports was my first solution. It seemed like a very easy thing to do, and it worked just fine. The only "problem" was that DECnet over asynch serial ports are slow, since they are limited to 9600 bps.
I might have run with that for at most a year before I wrote the bridge program to replace it.
I also needed physical serial ports all over the place. :-)
Yeah, it was a good first step though. Got the ball rolling. :)
-brian
On 2011-12-31 02.48, Steve Davidson wrote:
Johnny,
You are talking about the INFO.TXT file that is located in the default
DECnet directory.
Right. I even have one on MIM::
I just couldn't remember the name or format of the file, but figured someone else probably would. Little did I realized that others had already responded with the same information. :-)
The one "problem" with it, is that machines that aren't online also can't be queried for that file, so it might be a bit limited from that point of view.
Johnny
-Steve
-----Original Message-----
From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On
Behalf Of Johnny Billquist
Sent: Friday, December 30, 2011 8:46 PM
To: hecnet at Update.UU.SE
Subject: Re: [HECnet] Re: Topology 28-Dec-2011
Someone (I don't remember who) had a suggestion a year or two ago, that
machines should have a file in the default DECnet directory, with some
basic information about the node in a defined format, and the file with
a defined name, so that anyone could easily extract and figure out
information about atleast running nodes.
Maybe that idea could be expanded upon?
Seems very tedious for people to constantly mail updates... Although I
could of course take additional information as I'm already keeping a
nodename database in Datatrieve. No big deal to add additional
information if we'd like that... Then I can generate any kind of reports
based on this. As can others (if they can get remote Datatrieve to work
against RSX.)
Johnny
On 2011-12-30 15.59, hvlems at zonnet.nl wrote:
Rok, the spreadsheet that I distributed also has a sheet called
'hosts'.
I'm willing to expand that list though I have the basty feeling it's
going to take a lot more time than the project itself.
The information is not readily available thru NCP. Possibly NCL has a
relevant command but if so my ncl is *very* rusty and it seems I'm
pretty much the only phase V owner.
NCP gives limited information: the set exec node and sho exec (char)
commands only tell you the os name. For vms it tells you what platform.
The excel file contains this information for each node in a single
cell:
<address>-<name>-<hardware>-<os name>
Mark the hyphens: that way I can manipulate the csv file easily on vms
with f$element ;-)
So anyone who mails me node data in the format listed gets data
included in the file. What do you think?
Hans
-----Original Message-----
From: Rok Vidmar<rok.vidmar at gmail.com>
Sender: owner-hecnet at Update.UU.SE
Date: Fri, 30 Dec 2011 15:40:56
To:<hecnet at update.uu.se>
Reply-To: hecnet at Update.UU.SESubject: Re: [HECnet] Re: Topology
28-Dec-2011
Sorry, I'm confused. Is it your plan to keep a manually updated
list of
all of HECnet? Are you looking for data on all known nodes, even end
nodes
and ones that don't normally run 24x7? And how much data do you want
for
each node? OS? Hardware type? Geographic location?
That's fine with me and I'm not complaining in the least if that's
the
case! OTOH, I thought you were making a list of the active routing
nodes,
which is an entirely different matter.
Well, both. I generated list of active nodes and I believe it will
not be too
hard to manually update it from time to time.
I can add whatever interesting information and I can add any
number of
editors. Any volunteers? Hans?
--
Regards, Rok
On Dec 30, 2011, at 5:42, Johnny Billquist <bqt at softjar.se> wrote:
On 2011-12-30 02.44, Brian Hechinger wrote:
Agh, magica. Those were the good old days. :)
Magica is still around. I only wish I could turn her on more often. Cooling is the problem.
I really need to go dig my /70 out and start running it one of these days. That and my /34a. I don't currently have anywhere appropriate to run them, however. :(
Hum...? I think that it terminated on our VAX 8650.
That is the only thing I can remember on which I had serial ports to spare. Running Ultrix on the machine at the time.
Maybe that's it and I'm misremembering. I remember it ran Ultrix but remember it being MIPS for some reason.
That machine is also still around, but also powered off because of cooling. Big ECL machine. Lots of heat... But it's KRILLE (1.8).
It can boot VMS, Ultrix and NetBSD. Only got DECnet for VMS though.
Yeah, a shame it's too expensive to run. Nice machine.
-brian
Johnny,
You are talking about the INFO.TXT file that is located in the default
DECnet directory.
-Steve
-----Original Message-----
From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On
Behalf Of Johnny Billquist
Sent: Friday, December 30, 2011 8:46 PM
To: hecnet at Update.UU.SE
Subject: Re: [HECnet] Re: Topology 28-Dec-2011
Someone (I don't remember who) had a suggestion a year or two ago, that
machines should have a file in the default DECnet directory, with some
basic information about the node in a defined format, and the file with
a defined name, so that anyone could easily extract and figure out
information about atleast running nodes.
Maybe that idea could be expanded upon?
Seems very tedious for people to constantly mail updates... Although I
could of course take additional information as I'm already keeping a
nodename database in Datatrieve. No big deal to add additional
information if we'd like that... Then I can generate any kind of reports
based on this. As can others (if they can get remote Datatrieve to work
against RSX.)
Johnny
On 2011-12-30 15.59, hvlems at zonnet.nl wrote:
Rok, the spreadsheet that I distributed also has a sheet called
'hosts'.
I'm willing to expand that list though I have the basty feeling it's
going to take a lot more time than the project itself.
The information is not readily available thru NCP. Possibly NCL has a
relevant command but if so my ncl is *very* rusty and it seems I'm
pretty much the only phase V owner.
NCP gives limited information: the set exec node and sho exec (char)
commands only tell you the os name. For vms it tells you what platform.
The excel file contains this information for each node in a single
cell:
<address>-<name>-<hardware>-<os name>
Mark the hyphens: that way I can manipulate the csv file easily on vms
with f$element ;-)
So anyone who mails me node data in the format listed gets data
included in the file. What do you think?
Hans
-----Original Message-----
From: Rok Vidmar<rok.vidmar at gmail.com>
Sender: owner-hecnet at Update.UU.SE
Date: Fri, 30 Dec 2011 15:40:56
To:<hecnet at update.uu.se>
Reply-To: hecnet at Update.UU.SESubject: Re: [HECnet] Re: Topology
28-Dec-2011
Sorry, I'm confused. Is it your plan to keep a manually updated
list of
all of HECnet? Are you looking for data on all known nodes, even end
nodes
and ones that don't normally run 24x7? And how much data do you want
for
each node? OS? Hardware type? Geographic location?
That's fine with me and I'm not complaining in the least if that's
the
case! OTOH, I thought you were making a list of the active routing
nodes,
which is an entirely different matter.
Well, both. I generated list of active nodes and I believe it will
not be too
hard to manually update it from time to time.
I can add whatever interesting information and I can add any
number of
editors. Any volunteers? Hans?
--
Regards, Rok
On 2011-12-30 17.28, Gregg Levine wrote:
Incidentally Johnny Billquist, have we ever worked out a reasonable
method of assigning a dynamic IP address to be reachable by the
efforts of the network? I know it works for websites.....
Uh? Not sure what you mean?
Johnny
Someone (I don't remember who) had a suggestion a year or two ago, that machines should have a file in the default DECnet directory, with some basic information about the node in a defined format, and the file with a defined name, so that anyone could easily extract and figure out information about atleast running nodes.
Maybe that idea could be expanded upon?
Seems very tedious for people to constantly mail updates... Although I could of course take additional information as I'm already keeping a nodename database in Datatrieve. No big deal to add additional information if we'd like that... Then I can generate any kind of reports based on this. As can others (if they can get remote Datatrieve to work against RSX.)
Johnny
On 2011-12-30 15.59, hvlems at zonnet.nl wrote:
Rok, the spreadsheet that I distributed also has a sheet called 'hosts'.
I'm willing to expand that list though I have the basty feeling it's going to take a lot more time than the project itself.
The information is not readily available thru NCP. Possibly NCL has a relevant command but if so my ncl is *very* rusty and it seems I'm pretty much the only phase V owner.
NCP gives limited information: the set exec node and sho exec (char) commands only tell you the os name. For vms it tells you what platform.
The excel file contains this information for each node in a single cell:
<address>-<name>-<hardware>-<os name>
Mark the hyphens: that way I can manipulate the csv file easily on vms with f$element ;-)
So anyone who mails me node data in the format listed gets data included in the file. What do you think?
Hans
-----Original Message-----
From: Rok Vidmar<rok.vidmar at gmail.com>
Sender: owner-hecnet at Update.UU.SE
Date: Fri, 30 Dec 2011 15:40:56
To:<hecnet at update.uu.se>
Reply-To: hecnet at Update.UU.SESubject: Re: [HECnet] Re: Topology 28-Dec-2011
Sorry, I'm confused. Is it your plan to keep a manually updated list of
all of HECnet? Are you looking for data on all known nodes, even end nodes
and ones that don't normally run 24x7? And how much data do you want for
each node? OS? Hardware type? Geographic location?
That's fine with me and I'm not complaining in the least if that's the
case! OTOH, I thought you were making a list of the active routing nodes,
which is an entirely different matter.
Well, both. I generated list of active nodes and I believe it will not be too
hard to manually update it from time to time.
I can add whatever interesting information and I can add any number of
editors. Any volunteers? Hans?
--
Regards, Rok
