Zabbix gets strange data from nexus 5596UP

[dijix1990]

I have two nexus 5596UP are working in VPC and recently Zabbix starts to get strange data from one nexus 5596UP via snmp

There are no errors in the log

Maybe somebody know about it. I reboted it, but it didn't help

[marce1000]

 - Usually those traffic sparks are real but not  yet identified (which apps and or hosts causing it) , consider a netflow setup where then you can also analyze the traffic  (e.g.) ,

 M.

[dijix1990]

As I know the nexus 5596up can't use netflow. I use netflow with snmp on the cisco catalyst and on the every router. 

[dijix1990]

Maybe I bumped into a bug, but I don't access https://quickview.cloudapps.cisco.com/quickview/bug/CSCuy23473

[marce1000]

 - The bug is for CiscoPro Workgroup EtherSwitch Software , so unrelated , 

 M.

[dijix1990]

Does it have workaround? 

[MHM Cisco World]

this traffic can sync between two NSK SW.
NSK in vPC exchange Sync 

[dijix1990]

It happens with every interfaces it doesn't matter onboard port or extra module or fex

[dijix1990]

I reloaded again and it helped

[MHM Cisco World]

If happened again 

Show interface 

Check if unicast broadcast or multicast count increase rapidly

Yoh can have unknown unicast issue or broadcast strom in your case I dont think so but let us check.

[dijix1990]

I was watching. there was nothing special there. All nexuses work with snmp strangely . I mean, when I use a 1 minute interval to poll tx/rx traffic I see sawtooth graphs, independent of the polling system, zabbix/grafana/solarwinds. The survey works fine with an interval of 150 seconds