Months ago I wrote patch to print names for trap numbers from enterprise
traps.
Despite resending it at least once it seems to have been lost.
It might need some massaging due to the time lag.
Regards,
Andrew Hood
A distributed system is one in which the failure of a computer you
didn't even know existed can render your own computer unusable. --
Leslie Lamport, as quoted in CACM, June 1992
> -----Original Message-----
> From: Alex [mailto:alexcnt@xxxxxxxxx]
> Sent: Tuesday, 7 January 2003 03:39
> To: Ethereal Development
> Subject: [Ethereal-dev] Added the capability to decode SNMP OID from
> private MIB.
>
>
> Hi,
> I have added a new funtionality that permit to decode the
> Object Identifier
> on a SNMP packet with the definition declared on private MIB files.
>
> Actually Ethereal permit to show the name of the SNMP Object
> Identifier
> (OID) captured only if it's compiled with the UCD/NET-SNMP
> library, and the
> decoding of the OID happens only for standard MIBS and not
> for private MIB.
> These new features, that doesn't impact on the normal way as ethereal
> working, use a file that contains the OID and the name of the
> MIB variable.
> Below an example of the sintax of the file:
>
> 1 iso
> 1.3 iso.org
> 1.3.6 iso.org.dod
> 1.3.6.1 iso.org.dod.internet
> 1.3.6.1.4 iso.org.dod.internet.private.
> 1.3.6.1.4.1 iso.org.dod.internet.private.enterprises
>
> but the file can also contain a single line that has more associations
> between OID and variable mib name as reported below:
>
> 1.3.6.1.4.1 iso.org.dod.internet.private.enterprises
>
>
> All the associations between OID and the variable mib name
> present on the
> file will be stored on memory and for every SNMP packet will be show
> the decoding.
> It's possible to compile manually the file that contains the
> associations or
> use a script (I normally use a TCL script that uses a
> scotty-snmp library to
> compile my private MIB files and to produce the file used for the new
> functionality added on ethereal).
>
> I think that this new feature don't introduce regression at
> the normal way
> as ethereal working, the feature is running only if a file
> that contains the
> association is resent, otherwise the normal behavior is warranty.
>
> These new code, added on packet-snmp.c file, was tested on
> Linux RedHat
> 7.3/8.0 and with ethereal version 0.9.7 and 0.9.8.
>
> In attachment at this e-mail you can find a screenshot of how
> this feature
> working on ethereal with and without the support of NET-SNMP library.
>
> If somebody thinks that this feature is useful can reply at
> this mail to
> receive the source code and more explanations.
<<application/ms-tnef>>