If I submit this as an AUGMENT would it still mean that hrStorageAvailable
would appear grouped with the rest of the hrStorage table in an SNMP walk
or would it be a completely separate MIB?

I've read http://tools.ietf.org/html/rfc2578#page-29 quite a few times, but
I can't say that its any clearer now than it was the first read through.

On Wed, Sep 12, 2012 at 12:30 PM, Wes Hardaker <[email protected]> wrote:

> "Randy Presuhn" <[email protected]> writes:
>
> > Submit an internet draft.  The draft would contain a MIB module.
> > That MIB module would contain two definitions.  The first would
> > be for a "conceptual row" containing a single object (which you
> > defined as "hrStorageAvail" below) with an AUGMENTS clause
> > referring to the base conceptual row in the Host Resources MIB.
> > The second would be the definition you proposed below.
>
> <rant-with-no-real-expectations-of-change>
>
> You know, I know that's the right approach.  I've always known that.
> But...  I've always hated it.
>
> I've always failed to see why we can't add a new column to an existing
> table, which is all that is needed.  Simply republish the existing MIB
> with the new column.
>
> There are really only two downsides:
>
> 1) Badly written management code that is doing a full table walk and is
>    not looking at the output careful to realize it should discard some
>    things it doesn't care about being returned.  I honestly don't know
>    how much of this type of code exists in the world, where after the
>    last column in a table it can't handle extra data.  Maybe it's more
>    than I'd like.
>
> 2) It'd drop the existing DRAFT standard back to PROPOSED.  I actually
>    don't care about this too much.  We (the IETF) has already admitted
>    that we have issues with "levels".  In reality, we need only the new
>    parts to be PROPOSED and everything else remain DRAFT, but we can't
>    do that without doing the augment-in-a-new-draft-thing, which is
>    painful to the end user: IE, the people that matter.
>
> In netconf we now have the ability to augment in place (Yay!), but the
> output of netconf is less human readable in it's raw-ish form because of
> the namespace tag marking explosions (Boo!).
>
> Yes, I'm ranting about the known-failings of the SMI, which no matter
> how hard we try to move from it's still very very very important in the
> IETF world as new changes and new MIBs are being written constantly.
> But the failings of the SMI actually aren't the SMI at all, it's how we
> let ourselves use the SMI in publications "don't change it at all, even
> when it's safe to do so".
>
> [note that conformance statements aren't a problem, IMHO, you just need
> to add another and let the existing one remain.  A device can still be
> conformant to the old conformance requirements.]
>
> </rant-with-no-real-expectations-of-change>
>
> --
> Wes Hardaker
> My Pictures:  http://capturedonearth.com/
> My Thoughts:  http://pontifications.hardakers.net/
>  _______________________________________________
> OPSAWG mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/opsawg
>
_______________________________________________
OPSAWG mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/opsawg

Reply via email to