RE: snmpwalk & snmpget

I was going to suggest something like what Dave did...that this looks noncomforming.

To elaborate, it feels like the agent is being very picky (and apparently wrong) in the way it's parsing and interpreting OID's for snmpget (but it is doing the proper iteration for snmpnext PDU processing). I'm sure you've done something like this already but you might try, for the get case *specifically*, variations of trailing-zero/not-trailing-zero, a dot-following-the-last-suboid (i.e., ending the actual OID) vs. not providing same (a Novell agent had that bug a while back), a leading dot before the "1.3.6.1..." or not, etc.

More things to think about--what is the structure of this agent--is it a subagent in a system with something like Net-SNMP or SystemEdge of a Windows System agent handling the core non-enterprise MIBs and the SQL Server MIB is just a modular proxy or something? What I'm heading towards is that the idea of, for a scalar, doing

snmpget 1.3.6.1.<trailing OID scope>.0

should work for a get where the "trailing OID scope" gets you to something like a) the sysDescr, b) something in the Microsoft enterprise space that isn't SQL Server agent based (I don't know--something basic in a Windows Server system MIB, for example, that is in the "311" enterprise tree), and then ultimately to this SQL Server MIB scalar.

My thinking is, you may find that accepting a get with the "proper" .0 scalar final subOID will "break" somewhere in the MIB tree that is the handoff point to a subagent, get my drift? If so, you might be able to nail where that is and what it *would* accept as an alternate request OID for a get. It would be basically working "with" the bug in the subagent, but from your tone, it sounds like this kind of workaround would at least get you up and functioning in your effort a little bit.

That would be prima facie evidence of what Dave's suggesting, that there is something broken about this particular [sub]agent handling, and it wouldn't be surprising if it got "fixed" on you in the future, "breaking" whatever workaround you conjured using the process described here. Ultimately, if you have a developer support relationship with MS and the SQL Server group, this should be a very recreatable thing for them, and they could give you the skinny on why or whether they think this is correct behavior (it isn't) and if not, if they plan to fix it.

Microsoft accepts major credit cards for per-incident support calls. :) (in all seriousness, in my experience, they don't charge you if what you uncover in your support call is a bug in their product).

Regards,
Wayne Tackabury
Mirror Image Internet

-----Original Message-----
From: net-snmp-users-bounces@lists.sourceforge.net
[mailto:net-snmp-users-bounces@lists.sourceforge.net]On Behalf Of
net-snmp-users-request@lists.sourceforge.net
Sent: Friday, April 11, 2008 10:53 AM
To: net-snmp-users@lists.sourceforge.net
Subject: Net-snmp-users Digest, Vol 23, Issue 45


Send Net-snmp-users mailing list submissions to
net-snmp-users@lists.sourceforge.net

To subscribe or unsubscribe via the World Wide Web, visit
https://lists.sourceforge.net/lists/...net-snmp-users
or, via email, send a message with subject or body 'help' to
net-snmp-users-request@lists.sourceforge.net

You can reach the person managing the list at
net-snmp-users-owner@lists.sourceforge.net

When replying, please edit your Subject line so it is more specific
-------------
Date: Fri, 11 Apr 2008 13:51:28 +0000
From: "snmp girl" <snmp.girl@gmail.com>
Subject: Fwd: snmpwalk & snmpget
To: net-snmp-users@lists.sourceforge.net
Message-ID:
<b292483d0804110651q73dba995x66aa5daae40ac444@mail .gmail.com>
Content-Type: text/plain; charset="iso-8859-1"

I am using the microsoft agent
net-snmp installed is 5.4.1
what else do u need to know exactly???

---------- Forwarded message ----------
From: Dave Shield <D.T.Shield@liverpool.ac.uk>
Date: 11 avr. 2008 13:47
Subject: Re: snmpwalk & snmpget
To: snmp girl <snmp.girl@gmail.com>
Cc: net-snmp-users@lists.sourceforge.net

On 11/04/2008, snmp girl <snmp.girl@gmail.com> wrote:
> here what I get

[snip]

OK - that does indeed look wrong.
I'm inclined to suspect the code that implements this particular table.

Which SNMP agent are you using?
The Net-SNMP agent? The standard Microsoft agent?
Something else?

Dave
-------------- next part --------------
An HTML attachment was scrubbed...

------------------------------

Message: 4
Date: Fri, 11 Apr 2008 15:26:47 +0100
From: "Dave Shield" <D.T.Shield@liverpool.ac.uk>
Subject: Re: snmpwalk & snmpget
To: "snmp girl" <snmp.girl@gmail.com>
Cc: net-snmp-users@lists.sourceforge.net
Message-ID:
<c64ae3380804110726y40e8a221x6eb1284a9f8f891b@mail .gmail.com>
Content-Type: text/plain; charset=ISO-8859-1

On 11/04/2008, snmp girl <snmp.girl@gmail.com> wrote:
> I am using the microsoft agent

Then I suspect that this is something you may need to
take up with them. It very much feels like a problem
with the agent - not with the Net-SNMP tools.


> net-snmp installed is 5.4.1
> what else do u need to know exactly???

Do Microsoft provide any client-side SNMP tools?
It might be worth seeing if you can reproduce this
problem using their GET client application (if any).

But in any case, I think you probably need to
talk to Microsoft about this.

Dave



------------------------------

-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757...un.com/javaone
_______________________________________________
Net-snmp-users mailing list
Net-snmp-users@lists.sourceforge.net
Please see the following page to unsubscribe or change other options:
https://lists.sourceforge.net/lists/...net-snmp-users