For some reason this never hit the list - so Jason Read this one:
-----Original Message-----
From: Barr, Scott
Sent: Tuesday, October 11, 2005 10:52 AM
To: 'nv-l@lists.us.ibm.com'
Subject: RE: [nv-l] Duplicate interface that doesn't exist
Okay wait, I think I remember this
Is it possible you are using secondary addressing on the interface?
When you change the IP address on the interface you usually do a command
to add the IP address to the interface - that does not strictly remove
the old one. You can have multiple IP addresses on a router interface as
you will recall (secondary addressing and failover technologies that use
virtual interfaces come to mind).
The key I think is the fact that the "old" interface is operStatus down.
This is probably due to the fact that there is no route statement out of
the router for that interface (of course not, you changed the address
and thus the routing statements that go along with it). With no route
out the interface won't come up.
My gut says trying doing a no ip address command for the old address and
then demand poll the router.
Changing the old address should not change the interface to down but BE
CAREFUL.
-----Original Message-----
From: owner-nv-l@lists.us.ibm.com [mailto:owner-nv-l@lists.us.ibm.com]
On Behalf Of jason.duppong@thomson.com
Sent: Tuesday, October 11, 2005 10:27 AM
To: nv-l@lists.us.ibm.com
Subject: [nv-l] Duplicate interface that doesn't exist
Hi List,
Has anyone ever had a problem with Netview discovering duplicate fast
ethernet interfaces on nodes and the duplicated interfaces have the same
MAC address? A few weeks ago our WAN group readdressed the fast
ethernet side of a node. I was expecting that the readdressed IP would
simply be discovered by Netview during the next config scan and the
interface updated, but instead it added a new interface to the node,
duplicating the existing one and changing the IP. When the network
group told me about it I ended up removing the node and reloaded netmon.
We use the seed file exclusively for discovery and do not have
auto-discovery enabled. After the node was removed and Netview picked
up the device, both interfaces came back. I've since removed the device
several times and run ovtopofix hoping that it would correct something
goofy with my database, but it doesn't complain about anything being odd
with the node and every time the device is scanned Netview insists on
duplicating the interface. I double checked with the network group that
there is nothing odd with this device, all the interfaces we expect to
show up in Netview show up in addition to the one that Netview is
duplicating, the node is behaving normally and so is my Netview system.
A sample of the TopM Interface List is below:
"FastEthernet0/0 Up 10.222.144.1 255.255.255.128
0x000BBE0772E0 ethernet csmacd" <-- This exists
"FastEthernet0/0 Down 192.168.163.1 255.255.0.0
0x000BBE0772E0 ethernet csmacd" <-- This was the old IP
This problem is happening with both of my Netview servers and is
actually happening with two different nodes, both Cisco 2651s. This
leads me to believe it is related to our network setup, but I don't know
where to look anymore. If anyone has any suggestions they would be
greatly appreciated.
Thanks,
Jason
ProductName: "IBM Tivoli NetView"
OperatingSystem: "Linux"
Version.Release: "7.1.4"
Maintenance.Release: "7.1.4.0"
FixLevel: "7.1.4.3"
KanjiFixLevel: "None"
ClientFixLevel: "7.1.4.3"
|