Netstack IP Redirect Log Messages

[Ed Roche]

Good Morning!

We have come across a nuisance log entry with our new Nexus 9k top of rack switches. We constantly keep getting the following message:

"netstack[14753]: IP Redirect triggered on interface mgmt0"

We have a total of (8) Nexus 9K switches, (4) are setup in (2) different core pairs, (2) are setup as top of rack pair and the last (2) are a pair of distro switches. Each pair is setup to use vPC between them. Logging is set to default on all of them for the NETSTACK. This particular pair uses the MGMT interface as the keepalive for the vPC BUT we have one of the other pairs using the same config but no messages. The weird part is the packet that is triggering the message is destined for our domain controller. Also its a 30 minute cycle. So almost like a keepalive or heartbeat which for the vPC makes sense but not the part about it heading to the DC. At the top of the hour it goes one way then at the bottom it goes the other.

If anyone can shed some light on this it would much appreciated.

Thanks!

[sidshas03]

Good morning,

I noticed the log entry "IP Redirect triggered on interface mgmt0" on your Nexus 9k switches. This means ICMP redirects are being sent. You can fix this by disabling ICMP redirects on the mgmt0 interface with the command `no ip redirects`. Also, it might help to compare the configurations with the other switch pair that isn't having this issue.

Best,
Sid

[Ed Roche]

Hi Sid
I compared them already the older (4) switches have the same configuration for the vPC and then the (4) new ones are the same but only (2) of these new ones are kicking the message. I have tweaked out Solarwinds alert, so we stop getting the alert emails for now but need to get a change windows to test the change.

Thanks for your insight.
Ed

[Ed Roche]

Also meant to ask....is there a reason why traffic would be moving across the MGMT interface? Is it due to resource usage between those interfaces as opposed to the normal 1-48?