Symptom
An 8000 series distributed router may have traffic loss (dropped packets)
Conditions
This may occur, after an individual line card, that has a default route advertised via ebgp, is reloaded.
The symptoms may be seen with the following counters incrementing at a high rate:
CLI output where the statistics are incrementing at a high rate:
show controller npu stats traps-all instance all location | include L3_ROUTE_LOOKUP_FAILED
show cef drops location all | include "No route drops"
Workaround
Clear the RIB (router information database) via the following CLI command to re-sync the programming for IPv4/IPv6 routes:
clear bgp *
Further Problem Description
The following CLI statistics output will be increasing at a high rate on the impacted line card using
the following command on the line card having the problem:
show controller npu stats traps-all instance all location | i L3_ROUTE_LOOKUP_FAILED
Example if the slot 0 linecard (0/0/cpu0) is having the problem
8808#show controller npu stats traps-all instance all location 0/0/cpu0 | i L3_ROUTE_LOOKUP_FAILED
Mon Aug 16 11:03:54.808 EDT
L3_ROUTE_LOOKUP_FAILED 0 110 0x0 365 968 4468545
L3_ROUTE_LOOKUP_FAILED 1 110 0x0 365 330 1492613
8808#show controller npu stats traps-all instance all location 0/0/cpu0 | i L3_ROUTE_LOOKUP_FAILED
Mon Aug 16 11:03:54.808 EDT
L3_ROUTE_LOOKUP_FAILED 0 110 0x0 365 968 4479545 <-incrementing
L3_ROUTE_LOOKUP_FAILED 1 110 0x0 365 330 1509613 <-incrementing
RP/0/RP0/CPU0:8808#show cef drops location all | include "No route drops"
Mon Aug 16 11:16:41.733 EDT
No route drops packets : 6279806
No route drops packets : 675068681
No route drops packets : 413
RP/0/RP0/CPU0:8808#show cef drops location all | include "No route drops"
Mon Aug 16 11:16:45.148 EDT
No route drops packets : 6279806
No route drops packets : 684225892 <-incrementing
No route drops packets : 413
