| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 1797.1 | | MARVIN::COBB | Graham R. Cobb (Wide Area Comms.), REO2-G/H9, 830-3917 | Tue Nov 12 1991 12:49 | 16 |
| Re: .2. The Name attribute gives you the name of the node as currently
believed by the node itself.
A node may have zero, 1 or more than 1 name. The ideal is 1. My guess is
that your second node has 0 names (i.e. the node has never been told its
own name). In that case the Name attribute is set to the null fullname
which MCC appears not to be displaying correctly (the DNS architecture
defines that the null fullname should be displayed as 0:.).. If the node is
a WANrouter just issue a RENAME NODE xxxx NEW NAME xxxx command.
Note that the name you use to reference a node may or may not be the same
name that node believes it is called. For example the name you used might
be a softlink to the real name. Or the node name may have changed without
the node being told (with a RENAME directive).
Graham
|
| 1797.2 | My other questions | ANNECY::BONNIER | Jean-Luc ... EIC/T&N Annecy France | Wed Nov 13 1991 03:10 | 17 |
| Re: -1. Thanks Graham for your reply.
I didn't know that a node may have zero, 1 or more than 1 name.
Although it is not a WANrouter (it is a ULTRIX host with
"DECnet/OSI for Ultrix"), I just issued a
RENAME NODE xxxx NEW NAME xxxx command and now that works !!
BUT
1. what about the "self-creation" of the node entity as stated in the doc ?
2. what about managing nodes defined in different nameservers ?
3. what about the absence of DNA5 sections in the release notes ?
Thanks again.
Jean-Luc
|
| 1797.3 | | MARVIN::COBB | Graham R. Cobb (Wide Area Comms.), REO2-G/H9, 830-3917 | Fri Nov 15 1991 14:23 | 17 |
| If it is ULTRIX then it is recommended that you don't just use RENAME but
use their configuration script. The script issues the RENAME for you and
does several other things that are required.
> 1. what about the "self-creation" of the node entity as stated in the doc ?
What about it? Node entities are self-creating -- they have no CREATE
directive, they just create themselves when they boot for the first time.
What is the problem with that?
> 2. what about managing nodes defined in different nameservers ?
> 3. what about the absence of DNA5 sections in the release notes ?
No idea on those two -- I don't work on MCC.
Graham
|
| 1797.4 | What does CREATE mean ? | ANNECY::BONNIER | Jean-Luc ... EIC/T&N Annecy France | Thu Nov 28 1991 04:12 | 16 |
| Re:-1
> 1. what about the "self-creation" of the node entity as stated in the doc ?
What about it? Node entities are self-creating -- they have no CREATE
directive, they just create themselves when they boot for the first time.
What is the problem with that?
Graham,
The reason of my question is that :
In order to manage Phase V nodes from within DECmcc VMS 1.1, I had to register
them with the Iconic Map.
I understand there is no CREATE directive but when you say "create themselves",
what do you mean ? What means CREATE ?
JLuc
|
| 1797.5 | Create - as in LET THERE BE LIGHT! | BLUMON::SYLOR | Architect = Buzzword Generator | Thu Dec 05 1991 19:58 | 14 |
| Well, the Phase V NETMAN spec describes this in some detail. As you know, a
create causes an entity which didn't exist before to come into being - like the
VMS Create/DIR creates a directory. But there's no directive to "create" a node.
Instead, when a piece of hardware (CPU mostly) is first booted with the
right software, it will cause an entity to come into existance that we call
a node. Now this entity probably isn't yet registered in DNS (although it
could be), and DECmcc might not yet know about it( but it could*), still,
the entity exists!
Mark (once and probably future NETMAN architect)
--------------
* I think DECmcc still has the conceptual bug that it can't Register a node
that isn't reachable on the net, hope that got fixed.
|
| 1797.6 | | TOOK::SWIST | Jim Swist LKG2-2/T2 DTN 226-7102 | Fri Dec 06 1991 08:38 | 12 |
| Keep up with the implementors, Mr Architect...
Node reachability is no longer a criteria for registration (V1.2).
p.s. We are now working on a fix to allow unreachable entities to be
created. :-)
|
| 1797.7 | | MARVIN::COBB | Graham R. Cobb (Wide Area Comms.), REO2-G/H9, 830-3917 | Tue Dec 10 1991 08:14 | 14 |
| Re: .4. You may be confusing the notion of creating an entity with
registering an entity in the MCC map. These are completely independent: you
can register non-existent entities (in V1.2) and you can create entities
which are not registered.
When the hardware boots, it creates itself as a Node entity. That entity
now exists, on the network. It does the things a node does (see the NETMAN
architecture) and is manageable (at least locally, if not remotely).
At some other time (before or after) you register the node with MCC to make
it appear on the map. That registration has no effect on the node itself:
it purely affects MCC's databases.
Graham
|