On Tue, Oct 14, 2008 at 10:07:42AM +0200, Hofmann, Berthold wrote:
> we habe new cisco switches in our data center, Catalyst 4500 L3 software
> version 12.2(46) and see the following problem. Every 5 minutes we get this
> event:
> [...]
> We didn't change anything on the interfaces, and all switches do the same.
> Our older switches are not sending this messages, and we get no trap from the
> devices.
Unfortuantely, I don't have that particular chassis / version combo, but...
Is it possible that the snmp ifTable information being reported by the device
is incorrect? "Every 5 minutes" sounds like the default poll cycle. If
Spectrum is reading off bad information from the device, it may trigger the
interface reconfiguration event you're seeing (0x1001d).
For kicks,
- poll the system and ifTable mibs for this device from your command-line
and see what changes between five-minute polls. Ignore in/out octets and
other counters. Are the ifIndicies changing?
- via the Cisco cli, run "show snmp mib ifimib ifindex"; you should see a
list of your interfaces, do they look normal?
- via the Cisco cli, run "show run | i persist"... do you have "snmp-server
ifindex persist" enabled (this command became unsupported in this
version)?
- if you can sniff traffic, what does Spectrum poll before it fires off the
interface reconfiguration event?
--
yathrib
---
To unsubscribe from spectrum, send email to [EMAIL PROTECTED] with the body:
unsubscribe spectrum [EMAIL PROTECTED]
