SNMP Discovery Times Out but regular checks all work

Troubleshooting
1

CMK version: 2.5.0p8
OS version: Ubuntu 24.04

Error message: Services vanished: 72 (hr_cpu: 1, hr_fs: 42, if64: 18, juniper_cpu_util: 2, juniper_fru: 2, juniper_fru_fan: 2, juniper_mem: 1, juniper_temp: 1, pse_poe: 1, snmp_info: 1, uptime: 1), Vanished host labels: 3 (snmp_extended_info: 1, inv_lldp_cache: 1, ip_addresses: 1), [snmp] SNMP Error on 10.200.1.8: Timeout: No Response from 10.200.1.8 (Exit-Code: 1)CRIT, Automatic rediscovery currently not possible due to failing data source(s). Please run service discovery manuallyWARN, execution time 10.7 sec

Output of “cmk --debug -vvn hostname”: no error with this command. But adding -I to do a discovery returns several OIDs until it errors on one. The filesystem for JunOS which I don’t care to monitoring anyways.

Running ‘snmpbulkwalk -Cr20 -v2c -c XXXX -m “” -M “” -t 15.00 -r 2 -Cc -OQ -OU -On -Ot 10.200.1.2 .1.3.6.1.4.1.2636.3.1.13.1.8’
hr_fs: Fetching data (SNMP walk cache cleared)
Running ‘snmpbulkwalk -Cr20 -v2c -c XXXX -m “” -M “” -t 15.00 -r 2 -Cc -OQ -OU -On -Ot 10.200.1.2 .1.3.6.1.2.1.25.2.3.1.2’
Running ‘snmpbulkwalk -Cr20 -v2c -c XXXX -m “” -M “” -t 15.00 -r 2 -Cc -OQ -OU -On -Ot 10.200.1.2 .1.3.6.1.2.1.25.2.3.1.3’
ERROR: SNMP error: Timeout: No Response from 10.200.1.2

And running the below command on the same OID individually returns all of the values. I am not sure why it is timing out during the full discovery as all of the individual pieces seem to work fine.

snmpbulkwalk -v2c -c msd_public 10.200.1.2 .1.3.6.1.2.1.25.2.3.1.3

2

If you get the error on this snmp section

you should now create a rule to disable this snmp section for this device.
Setup → Disable or enabled sections (SNMP)
After the rule creation you can check on the command line if the scan runs without error or if there are more sections you need to disable.

1 Like
3

That is fantastic. My Juniper switches are always slow via SNMP checks and I could never find how to stop a check on an OID. I now understand my debug log when doing an inventory and how to stop checks I don’t care about instead of just disabling the services after they are discovered. 10 out 10, thank you!