N7K hardware rate-limiter

[Geir Sand-Strand]

Hi

We see some rate-limiter dropped packets with the show hardware rate-limiter command.  L3 ttl and L3 glean are showing dropped packets.

 R-L Class           Config        Allowed         Dropped            Total
 +------------------+--------+---------------+---------------+-----------------+

  L3 ttl                   500         1911655            5266           1916921
  L3 glean              100         1412552            3393           1415945

Since we are having some issues with Citrix Netscaler failover, we believe this could be due to dropped arp.

Changing the L3 ttl and glean value, will this impact trafficflow or could this command be done while production traffic is flowing in the corenetwork?

I can see that the values can be set between 0 and 30000.  What value would be "safe" to set? The n7k cpu have an average of 10-15% utilization.

thx

Geir

[Georg Pauwen]

Hello,

rather than changing the values you might want to use the ip glean throttle feature, which should get rid of the drops:

N7K(config)#hardware ip glean throttle

https://www.cisco.com/c/en/us/support/docs/switches/nexus-7000-series-switches/200677-Nexus-7000-Understanding-hardware-ip-g.html

[Geir Sand-Strand]

Thx for the quick reply.

The hardware ip glean throttle is a "safe" command to run while networktraffic is flowing in our corenetwork? Or should we do it in a maintenance window?

Geir

[Georg Pauwen]

Hello,

the feature installs a drop index, depending on what you are running on your network, I would, to be on the safe side, make the change after hours..

[Geir Sand-Strand]

Did run the command:

N7K(config)#hardware ip glean throttle

But it didn't seem to solve the issue.  So we need to look more into this.

Read that as a first step could be disable ip redirects on your svi Interfaces.  Think this is already done, but will double check that.

Geir