Nexus 5672UP packet loss - question on ASIC load balancing

[spfister336]

We have a pair of Nexus 5672UPs that occasionally experience packet losses, especially when backups are taking place to devices connected to these switches. I've made a list of what devices are using what ASIC. Is there any easy way to see if any one ASIC is being overloaded? We're currently only using the first 48 ports, not the 6 ports at the right.

[Georg Pauwen]

Hello,

in the document linked below, go to the section:

Check for Forwarding Manager (FwM) Drops

https://www.cisco.com/c/en/us/support/docs/switches/nexus-5000-series-switches/116172-technote-product-00.html#anc3

[spfister336]

Thank you for the link. I'm having some trouble figuring out some of the output. For example:

vmw3# show platform fwm info asic-errors 1

Printing non zero Carmel error registers - 48 bits:
BIG_DROP_INGRESS_UC_PC_DROP: res0 = 39 res1 = 0 [19]
BIG_DROP_HIT_DROP_PORT_MAP_IDX: res0 = 35 res1 = 0 [53]
BIG_DROP_SRC_MASK_TO_NULL: res0 = 27447 res1 = 0 [54]
BIG_DROP_SRC_VLAN_MBR: res0 = 23511 res1 = 0 [59]
BIG_DROP_EGRESS_VLAN_MBR: res0 = 38 res1 = 0 [79]

Printing non zero Carmel error registers - 32 bits:
vmw3# show platform fwm info asic-errors 2

Printing non zero Carmel error registers - 48 bits:
BIG_DROP_INGRESS_UC_PC_DROP: res0 = 19 res1 = 0 [19]
BIG_DROP_HIT_DROP_PORT_MAP_IDX: res0 = 17 res1 = 0 [53]
BIG_DROP_SRC_MASK_TO_NULL: res0 = 21559 res1 = 0 [54]
BIG_DROP_SRC_VLAN_MBR: res0 = 41538 res1 = 0 [59]
BIG_DROP_EGRESS_VLAN_MBR: res0 = 29 res1 = 0 [79]