Cisco Prime Infrastructure 2.1: SNMP Connectivity Failed

[albertofdez]

Hi,

I have discharged all my Cisco devices within Cisco Prime and after a few days and when everything worked correctly one of the switches is displayed with the SNMP error Connectivity Failed, nothing has been changed.

I deleted and i have returned to create the object but it continues to be the same, I've also tried to create another SNMP community still receiving the same error.

This is the configuration of snmp in the switch WS-C3850-48P with version 03.02.02.SE

snmp-server community "community" RW
snmp-server location "location"
snmp-server contact "contact"
snmp-server host 10.180.5.22 version 2c "community"

Best regards.

[AFROJ AHMAD]

Hi ,

SNMP config is fine. Is there any access-list applicable here ?

try to do the SNMPWALK on this device  and enabled the SNMP debug on the device (debug snmp packets) and share the logs.

Thanks-

Afroz

[albertofdez]

Hi,

There is no access-list nor have we made any changes to the configuration and before everything worked correctly.

This is the result of the debug snmp packets

Jun 16 09:10:13.914: SNMP: Packet received via UDP from 10.180.5.22 on Vlan80
Jun 16 09:10:13.914: %SNMP-3-INPUT_QFULL_ERR: Packet dropped due to input queue full

Best regards.

[AFROJ AHMAD · Accepted solution]

Hi ,

follow the below steps:

first copy and save the existing SNMP config  then issue the below command :

conf t# no snmp-server

now enter any one of the SNMP commands that you had

conf t# snmp-server community "community" RW

Now ,check the issue., I mean try to add the device back in Prime .

Also check the status of cpu utilization  , is CPU utilization normal ?

Thanks-

Afroz

***Ratings Encourages Contributors ***

[albertofdez]

Thanks

[albertofdez]

Hi,

Has become the same thing is happening again. The error that shows is the same as before, I have removed the SNMP community, I have become to remove and everything is ok, but at the time I see the error.

Do not think that it is a bug of IOS because I have the same team managed by the PI, with the same version and i have no problems.

Best regards.

[AFROJ AHMAD]

Hi ,

check if you are getting these errors in "show log" again:

 %SNMP-3-INPUT_QFULL_ERR: Packet dropped due to input queue full

If yes , then either a lot of SNMP polling happening on the device

or

CPU utilization is going high sometime , being SNMP as a least priority process , you will see this issue. check for "show proc cpu history"

It is a device side issue definitely.

I would suggest to apply a Access-list on the SNMP community string allowing only the

valid NMS to poll the device

Thanks-

Afroz

***Ratings Encourages Contributors ***