Exceeded the configured timeout of 15 minutes in SNMPCOLLECTOR

[Miller Echagarreta]

Hi community...

I have a network with 1000 devices in monitoring with snmpcollector and snmpv 2 and 3.
There are 40 devices that it's not possible discovery from snmpcollector. I see that these devices response to ping. For example...

C:\Users\ca>ping 10.200.15.193

Pinging 10.200.15.193 with 32 bytes of data:
Reply from 10.200.15.193: bytes=32 time=2138ms TTL=251
Reply from 10.200.15.193: bytes=32 time=2167ms TTL=251
Reply from 10.200.15.193: bytes=32 time=2280ms TTL=251
Reply from 10.200.15.193: bytes=32 time=1926ms TTL=251

Ping statistics for 10.200.15.193:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 1926ms, Maximum = 2280ms, Average = 2127ms

The are snmp credentials are working.

But SNMPCOLLECTOR indicate this message on this 40 devices... 

Discovery Failed - Discovery of device BOG_MED_STO386.local.com.co exceeded the configured timeout of 15 minutes

I try testing change the SNMP Timeout in Milliseconds = 5000 with SNMP Retries = 3, another configuration  SNMP Timeout in Milliseconds = 10000 with SNMP Retries = 2, in other device with SNMP Timeout in Milliseconds = 30000 with SNMP Retries = 2 but not work. For example...



Why happen this?
What I need configure in snmpcollector for this working?
Maybe it's a problem of OS for this 40 devices?

The devices are Cisco 1905k 9,  Cisco 1721,  Cisco Cat 295024

Any idea???

[David Michel]

First a side note. The number of metrics being monitored is of more value than the number of devices. 
Monitor Polling Statistics
The URL to access the servlet is: <IPaddress>:9715/metrics.
Use the IP address for the robot on which the snmpcollector probe is installed.

The IOS could be a contributing factor. Check to see if you have others of the same models with different revs that are working fine, as well as with Cisco support. 

For the profile, test it with a lower timeout say 5000 with 7 retries. 

As a sanity check remote into the system hosting snmpcollector and run a sapwalk, and see if that gets a response back right away or also delays. 
The tool is available at the device support site. 
https://devicesupport.broadcom.com/

[Miller Echagarreta]

Hi David, how are you...?

I followed your recommendation, and I tested with a device.

change the the snmp timeout, the retries and the message is =
too i tested with sapwlak -----> Discovery Failed - Discovery of device 10.000.00.81 exceeded the configured timeout of 15 minutes

The sapwalk end with error. This started 11:13 am and the end to 12:37 pm.




The admin of the host reviewed th OS with another host of the same model that if is on monitoring without problem and the OS is the same.

I can think that the sapwalk tells me that the problem is the device that at some point is locked to deliver snmp data, but how to tell the administrator what to do to solve it. Can you think of anything else I can do...?

[David Michel]

How about asking the device admin to run a sapwalk so he can see for himself. 
if not perhaps: 
Install wireshark on the snmpcollector hub, start the packet capture, run the sapwalk. 
When it stalls, stop the packet capture. 
Add a filter 'ip.addr eq x.x.x.x' to show only traffic for this device, save, and provide to the device admin. 
The goal is to show that the network level communication between systems is good, and so it needs to be addressed at the device end.

[Miller Echagarreta]

Excellent idea David, I will review your suggestion. Thanks...!!!

[Miller Echagarreta]

Hi @David Michel how are you?​ I have a question, i don't if you can help me with my doubth.
an device with problems of discovery where the bandwidth is very low, the administrator can configure snmp so that it only delivers availability, reachability and interfaces (bitsIN bistOUT)?

[David Michel]

Check out snmpget, that will probably work better for that device than snmpcollector.