Mike.
Most often, the sysObjectID.0 contains the enterprise OID, or
a derivation of this OID.
Have a look at:
ftp://ftp.isi.edu/in-notes/iana/assignments/enterprise-numbers
--
o Lecteurs francophones: j'ai un peu de mal a utiliser les accents
avec mon clavier QWERTY, alors ne m'en veuillez pas si je les oublie.
o Le meilleur produit du monde : http://www.openmaster.com
Any ideas?
Thanx in advance.
Mike.
Olivier Miakinen wrote in message <35BDCB11...@bull.net>...
Actually, sysObjectID should do just that. If not, either the implementation
or the MIB is incorrect.
Arnoud.
There you will have to rely on the manufacturers mibs. For Cisco, see
ftp://ftp.cisco.com/pub/mibs/v?/CISCO-PRODUCT-MIB.my
it lists most Cisco product (but strangely not Catalyst 5k's)
/Niels
--
Niels Baggesen, UNI-C, Olof Palmes Alle 38, DK-8200 Aarhus N, Denmark
Email: Niels.B...@uni-c.dk Tel: +45 89 37 66 69 Fax: +45 89 37 66 77
--- Never underestimate the bandwidth of a CD flying through the lab ---
Arnoud Zwemmer wrote in message <6prtd8$t7$1...@news2.xs4all.nl>...
>Mike Diehl wrote in message <6prc8p$d07$1...@news.rt66.com>...
>>This is true; this information is also in (I believe) rfc1700.
>>Unfortunately, it will only identify the manufacturer, or owner of the
>>namespace. I'm actually looking for a means of using this variable to
>>distinguish, say, a Cisco Catalyst 5000 from Cisco Catalyst 5500.
>
>
You'd have to collect all vendor-specific MIBs from vendors you're likely to
encounter equipment of. If you compile all these into your NMS, your NMS
will know, by getting sysObjectID, what type of equipment it sees.
Good luck,
-= Arnoud =-
Mike
Arnoud Zwemmer wrote in message <6q3r9i$nee$1...@news2.xs4all.nl>...
>Mike Diehl wrote in message <6q3kf3$gu$1...@news.rt66.com>...
>>I guess what I'm looking for is a database that says "if you get "X" ,back
I don't know if you have the oid_to_type or oid_to_sym files from Netview or
Openview - these are a mapping of enterprise OID to icons & snmp
characteristics used by NV/OV. They include a large number of device OID's and
would be a good starting point - I will email you a copy, including recent
additions to the file that I have made.
A word of warning as you undertake this massive task - in many cases the
enterprise OID is tied up with the firmware or O/S version, not the hardware -
for example I have two identical Cisco Routers with different versions of IOS
and different ent.OIDs; another example - NT 3.51 & NT 4 have the same ent.OID
- until you add the service packs.
The point being that the list you want to compile is very larg, and very
dynamic.
Good luck.
Adam
In article <6qo7hq$hno$1...@news.rt66.com>,
"Mike Diehl" <auto...@iname.com> wrote:
> Ok, so I'd have to collect all of this data. Anyone interested in
> contributing? I'd make the results widely available. What do you think?
>
> Mike
-----== Posted via Deja News, The Leader in Internet Discussion ==-----
http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
Please note that when trying to obtain a list of values returned for
sysObjectID that there will be problems caused the the definition
of the object.
This is because some SNMP agent designers interpreted the object
is identifying hardware, others interpreted the object as identifying
software, and other interpreted it as identifying both (and may have
encoded h/w and s/w in the value).
And there may be a few agent designers that "didn't do the right thing"
and
1) used the OID value from the SNMP agent toolkit designer,
2) used an OID that was not assigned to them, or
3) didn't change the OID value for new/different h/w or s/w.
Putting together a list will show places where there is a
problem, if any.