Hi Emile/PCE group members,
Thanks for this message. I have gone through this thread and the existing
MIBs. I have listed my comments/observations below:
1. In general, we do not include "STD" for a draft MIB. Only
when they are about to become RFC, we call them STD.
Before becoming an RFC we have "DRAFT" for the MIBs.
This way if some companies want to implement the draft versions
and later want to upgrade to standard versions do not have a name
clash with the OIDs.
2. I agree with most of your decomposition.
However, I want to propose :
PCEP-DRAFT-MIB draft-ietf-pcep-mib
instead of
PCE-COMP-STD-MIB draft-ietf-pce-comp-mib
This will give a clear idea to the user on what the MIB contains.
So the new decomposition looks like:
MIB module WG draft name
================================================
PCE-DISC-DRAFT-MIB draft-ietf-pce-disc-mib
-- PCE discovery MIB.
PCEP-DRAFT-MIB draft-ietf-pcep-mib
-- PCE communication protocol MIB
PCE-PCC-DRAFT-MIB draft-ietf-pce-pcc-mib
-- Path Computation Client MIB
PCE-TC-DRAFT-MIB draft-ietf-pce-tc-mib
-- PCE Textual conventions MIB.
3. I see that the SMIv2 structures of the existing MIB modules
need to be fixed and SMIC utility must be run. I can take care of that.
4. I will also start working on the PCEP-MIB for which you will
be my co-author.
Please let me know if you have any more comments/observations? I
intend to start working on these soon.
Regards,
Kiran.
----- Original Message -----
From: "STEPHAN Emile RD-CORE-LAN" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>; "A S Kiran Koushik" <[EMAIL PROTECTED]>;
<[EMAIL PROTECTED]>
Cc: "LE ROUX Jean-Louis RD-CORE-LAN" <[EMAIL PROTECTED]>
Sent: Wednesday, November 29, 2006 10:46 AM
Subject: PCE MIB modules
Hi Kiran, Adrian and JP
I am really happy that Kiran join us. I am sure that we are going to do
excellent works together... and that there is enough work for 2 people.
So let's start.
In July we identify the following MIB modules decomposition.
MIB module WG draft name
================================================
PCE-DISC-STD-MIB draft-ietf-pce-disc-mib
PCE-COMP-STD-MIB draft-ietf-pce-comp-mib
PCE-PCC-STD-MIB draft-ietf-pce-pcc-mib
PCE-TC-STD-MIB draft-ietf-pce-tc-mib
draft-ietf-pce-disc-mib and draft-ietf-pce-tc-mib are already in progress.
As JP spoke recently of a "PCEP MIB", I want to be sure that we still agree on
this decomposition?
Then I propose to draft the abstract of each document conjointly with the
chairs.
Finally, I think that we may identify the drafts and the sections giving
directions to write each MIB module.
At this point, each of us will share the minimal understanding before writing
any SMI line.
Regards
Emile
-----Message d'origine-----
De: Adrian Farrel [mailto:[EMAIL PROTECTED]
Envoy: vendredi 28 juillet 2006 14:08
: STEPHAN Emile RD-CORE-LAN
Cc: LE ROUX Jean-Louis RD-CORE-LAN; [EMAIL PROTECTED]
Objet: Re: PCE MIB modules names
Good work.
Thanks.
Adrian
----- Original Message -----
From: "STEPHAN Emile RD-CORE-LAN" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Cc: "LE ROUX Jean-Louis RD-CORE-LAN" <[EMAIL PROTECTED]>;
<[EMAIL PROTECTED]>
Sent: Friday, July 28, 2006 11:02 AM
Subject: RE: PCE MIB modules names
Hi Adrian,
I think we got it:
MIB module WG draft name
================================================
PCE-DISC-STD-MIB draft-ietf-pce-disc-mib
PCE-COMP-STD-MIB draft-ietf-pce-comp-mib
PCE-PCC-STD-MIB draft-ietf-pce-pcc-mib
PCE-TC-STD-MIB draft-ietf-pce-tc-mib
According to that I will firstly review and rename my current draft as
draft-stephan-pce-disc-mib-00.txt. Then I will create
draft-stephan-pce-tc-mib-00.txt. Finally I will post both of them.
Regards
Emile
-----Message d'origine-----
De : Adrian Farrel [mailto:[EMAIL PROTECTED]
Envoy : mercredi 26 juillet 2006 18:04
: STEPHAN Emile RD-CORE-LAN; [EMAIL PROTECTED]
Cc : LE ROUX Jean-Louis RD-CORE-LAN
Objet : Re: PCE MIB modules names
Hi Emile,
> During the meeting we decided to move the TCs defined in
> draft-stephan-pcc-pcedp-mib-01.txt in a dedicated MIB
> module to permit them to be shared by any (PCE) MIB
> modules;
Yes please.
> Prior to do that I would like to determine the names of
> the PCE MIB modules and the name of the corresponding
> drafts. The aim is to avoid collisions in the names of the
> MIB modules and drafts, renaming, misunderstanding...
> and annoying discussions in the ML for people not
> interested by MIB aspects.
This is very sensible, to avoid the collisions and any thrashing.
It is not so important to avoid the mailing list. I agree that not many
people are interetsed in MIB modules, but it is important to make sure
that
those who might be interested feel that there is a discussion going on
that
they can contribute to. Also, it is nice to show some interactions about
the
I-D becuase it helps to build evidence of support for the work.
> I identify 3 potential modules plus the TC one:
>1- A module to monitor from within a PCC (a PCE may
> includes a PCC), the PCEs discovered by this PCC or
> PCE (the intend of pcc-pcedp-mib);
Right. So this controls and monitors the discovery process, and reports on
discovered PCEs.
Presumably it also allows you to configure the location and capabilities
of
remote PCEs.
> 2- A module to monitor from within a PCE, path
> computation activities of this PCE;
Yes. I assume this includes the PCEP activities.
> 3- A module to monitor from within a PCC (a
> PCE may includes a PCC), path computation
> requested by this PCC or PCE;
Yes.
> I propose the 4 following names:
>
> MIB module WG draft name
> ========================================
> PCC-PCEDP-MIB draft-ietf-pce-pcc-pcedp-mib
> PCE-PATHCOMP-MIB draft-ietf-pce-pce-pathcomp-mib
> PCC-PATHCOMP-MIB draft-ietf-pce-pcc-pathcomp-mib
> PCE-TC-MIB draft-ietf-pce-tc-mib
Now, I have always had some trouble understanding how MIB modules are
supposed to be named, but from experience, it seems that we need...
xxx-yyy-STD-MIB
...where
xxx is the field of applicability (e.g. MPLS, GMPLS, PCE, etc.)
yyy is the thing being modeled.
STD means "standard"
So this would give us:
PCE-DISC-STD-MIB
PCE-COMP-STD-MIB
PCE-PCC-STD-MIB
PCE-TC-STD-MIB
The document names are fine as you feel appropriate.
Note that you are allowed to put multiple MIB modules in the same I-D, but
actually four documents is probably easier for everyone.
> The intent of the proposal above is only to start the discussion.
Yes. And I very much appreciate it. Thanks for driving this.
> We may use shorter names like in ccamp or mpls WG.
> What about PCC-DP-MIB, PCE-PC-MIB PCC-PC-MIB
> and PCE-TC-MIB?
My suggestions, above, are to keep the names looking a little less the
same!
Alternatives for the middle two of PCE-PCE-STD-MIB and PCE-PCC-STD-MIB
just
looked too confusing!
Cheers,
Adrian
