Hello,
It happened to 2 similar 2960X stacks (not at the same time).
These stacks are queried at regular basis by an MRTG server through SNMP.
Suddenly the stack stops responding to snmp queries. I succeeded in activating snmp debug once, so I saw incoming requests logged by the switch, but no packet has been sent in reply (I wonder if packet treatment is not achieved after retry query arrives). So, some time later appears the following message:
Dec 18 18:53:29 cisco-1-10 -: 2233966: 2233872: Dec 18 17:53:28: %SNMP-3-INPUT_QFULL_ERR: Packet dropped due to input
Dec 18 18:53:29 cisco-1-10 -: 2233967: queue full
The situation escalates to multiple panics for HRPC:
Dec 19 04:26:37 cisco-1-10 -: 2263213: 2248879: Dec 19 03:26:36: %SYS-2-MALLOCFAIL: Memory allocation of 42544 bytes failed from 0x290F6CC, alignment 0
Dec 19 04:26:37 cisco-1-10 -: 2263214: Pool: Processor Free: 135964 Cause: Memory fragmentation
Dec 19 04:26:37 cisco-1-10 -: 2263215: Alternate Pool: None Free: 0 Cause: No Alternate pool
Dec 19 04:26:37 cisco-1-10 -: 2263216: -Process= "HRPC qos request handler", ipl= 0, pid= 172
Dec 19 04:26:37 cisco-1-10 -: 2263217: -Traceback= 54E64Cz 21C4FA4z 21CDC7Cz 2FF2D50z 290F6D0z 2E32A68z 2E2E764z
And finally one of the switches reloads automatically.
I have to mention that once the snmp stops responding (or even before ?) the following command is blocked until the switch reloads automatically:
cisco-1-10#show int Gi1/0/49
Message from syslogd@cisco-1-10 at Dec 19 04:32:21 ...
Dec 19 04:32:21 cisco-1-10 -: 2263574: 2248974: Dec 19 03:32:20: %ENTROPY-0-ENTROPY_ERROR: Unable to collect sufficient entropy
packet_write_wait: Connection to 10.12.11.10: Broken pipe
Here is a part of the inventory command:
NAME: "1", DESCR: "WS-C2960X-48TD-L"
PID: WS-C2960X-48TD-L , VID: V05 , SN: FOCXXXXXXXX
NAME: "Switch 1 - FlexStackPlus Module", DESCR: "Stacking Module"
PID: C2960X-STACK , VID: V01 , SN: FOCXXXXXXXX
NAME: "GigabitEthernet1/0/49", DESCR: "1000BaseSX SFP"
PID: , VID: , SN: XXXXXX
Switch 01
---------
Switch Uptime : 3 days, 6 hours, 3 minutes
Base ethernet MAC Address : 04:2A:E2:DC:FF:00
Motherboard assembly number : 73-16688-03
Power supply part number : 341-0537-02
Motherboard serial number : FOCXXXXXXXX
Power supply serial number : DCB194981P8
Model revision number : M0
Motherboard revision number : B0
Model number : WS-C2960X-48TD-L
Daughterboard assembly number : 73-14200-03
Daughterboard serial number : FOCXXXXXXXX
System serial number : FOCXXXXXXXX
Top assembly part number : 68-100466-01
Top assembly revision number : B0
Version ID : V05
CLEI Code Number : CMMNH00ARD
Daughterboard revision number : A0
Here is the software version and the global uptime (only the faulty switch has reloaded 3 days ago) :
cisco-1-10 uptime is 1 year, 49 weeks, 6 days, 20 hours, 50 minutes
System returned to ROM by power-on
System restarted at 12:45:43 UTC Fri Jan 6 2017
System image file is "flash:/c2960x-universalk9-mz.152-2.E3/c2960x-universalk9-mz.152-2.E3.bin"
Last reload reason: Reload command
Thanks.
