Interleaved below...
On 2/11/08, Dave Shield <[EMAIL PROTECTED]> wrote:
>
> On 11/02/2008, Larry Dickson <[EMAIL PROTECTED]> wrote:
> > Response to getnext at netSnmp.1.3 is timeout.
>
> OK - what about
> GETNEXT netSnmp.1.5
> GETNEXT netSnmp.1.7
> GETNEXT netSnmp.1.9
>
> What I'm trying to determine here is exactly *where* the blocking module
> is located. If netSnmp.1.3 fails, and netSnmp.2 works, then it must be
> somewhere between these two points.
Getnext results for netSnmp.x
1.4 nsCacheDefaultTimeout.0
1.5 same
1.6 nsDebugEnabled.0
1.7 same
1.8 enterprises.22782.1.2.1.1.0 = Timeticks (start of our subtree)
The queries listed above are designed to test particular MIB modules.
>
>
>
> > Response to get at
> > netSnmp.1.3, netSnmp.1.3.1, etc, is noSuchName
>
> That makes sense.
> Those OIDs mark the top of the appropriate subtree, so wouldn't
> have a value associated with them. It's only if you tried something
> like
> snmpget .... netSnmp.1.3.2.1.0
>
> that I'd expect you to get an answer (nsExtendNumEntries.0)
netSnmp.1.3.2.1 gave noSuchName Failed object: nsExtendNumEntries and
netSnmp.1.3.2.1.0 timed out.
> My guess is that whatever comes immediately
> > after netSnmp.1.2.1.1.6.0.12.1.3.6.1.6.3.16.1.5.2.1.6.127
> > is where the damage is.
>
> And that's what I'm trying to pinpoint.
>
>
>
> > But my question is also more algorithmic: How is it (the agent/subagent
> > combination) supposed to do a getnext that jumps out of the subagent's
> > subtree?
>
> The subtree is given an OID, and returns the next value that it knows
> about.
What if it doesn't know what comes next, i.e. the subagent was asked
"getnext" for the (in lexicographic order) last leaf in its subtree for
which a value exists?
Larry Dickson
The master agent looks at this result, and decides whether the OID falls
> within
> the block registered by the subagent. If it does, then it returns that
> result
> to the original client application, and everyone is happy.
>
> If the subagent's answer comes *after* the end of the registered block,
> then
> the master agent looks to see whether there's another MIB module which
> has registered for a subtree before the returned OID. If there is, then
> the
> master agent will send the query to that module instead.
> It'll keep doing this until either it gets an acceptable answer, or it
> runs
> out of MIB modules to ask.
>
> Dave
>
-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Net-snmp-coders mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/net-snmp-coders
