Symptom
%SNMP-3-INPUT_QFULL_ERR: Packet dropped due to input queue full
Conditions
This problem is more likely to occur as the number of interfaces installed in the device increases.
PGHS-7609-1#sh proc cpu sorted | exc 0.00
CPU utilization for five seconds: 98%/8%; one minute: 98%; five minutes: 98%
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
654 32068748 128634330 249 46.47% 15.54% 24.56% 0 PDU DISPATCHER
521 9186152 2476029 3710 15.26% 12.93% 14.77% 0 Syslog Traps
249 490507980 2328850353 0 5.89% 5.47% 5.47% 0 IP Input
616 2028236 677408 2994 3.31% 3.75% 3.49% 0 Syslog
643 9024976 619886884 14 2.82% 8.40% 8.38% 0 HSRP Common
9 54320264 16496919 3292 2.48% 0.73% 0.81% 0 Check heaps
14 85883340 101112080 849 1.65% 1.52% 1.56% 0 ARP Input
571 209844456 787643151 266 0.74% 0.70% 0.68% 0 Port manager per
472 157302156 830298477 189 0.74% 0.62% 0.57% 0 LFDp Input Proc
Workaround
NOTE: this workaround will prevent NMS servers from polling information about interface statistics, so use with caution
snmp-server view cutdown iso included
snmp-server view cutdown 1.3.6.1.2.1.2.2.1 excluded
snmp-server community [community name] view cutdown RO 98
Perform a warm restart of snmp engine after the above applied workaround
Further Problem Description
Also increased the queue-length to 1000 ( Maximum extendible up to 5000) using command:
snmp-server queue-length 1000 -> Since device is being heavily polled and there are many traps that was configured on the device
