Attempting to configure WFA SNMP messages to Telegraf (InfluxDB/Grafana)

ChadPruden · ‎2017-04-11

I am struggling to get SNMP messages sending to a Telegraf instance.

1) I've opened appropriate network ports (outgoing/UDP161 --> incoming/UDP162)

2) Read the WFA admin guide section on SNMP

3) Enabled and configured SNMP within WFA (see screenshot)

4) Configured telegraf.conf and running a test process

Is anyone else successfully sending SNMP traps? Anyone sending to Telegraf/InfluxDB/Grafana?

telegraf.conf

[[inputs.snmp]]
agents = [ "10.65.163.71:161" ] ##WFA host
## Timeout for each SNMP query.
timeout = "60s"
## Number of retries to attempt within timeout.
retries = 5
## SNMP version, values can be 1, 2, or 3
version = 2

## SNMP community string.
community = "public"

## The GETBULK max-repetitions parameter
max_repetitions = 10

## measurement name
name = "system"
[[inputs.snmp.field]]
name = "wfa"
oid = "1.3.6.1.4.1.789.1.1.12.0"

[[outputs.kafka]]

....

sinhaa · ‎2017-04-12

@ChadPruden

[[inputs.snmp]]
agents = [ "10.65.163.71:161" ] ##WFA host
## Timeout for each SNMP query.

...

The ports are different. In the Image your WFA is configured for sending SNMP on Port 162.

If this post resolved your issue, help others by selecting ACCEPT AS SOLUTION or adding a KUDO.

sivakumar_sekar · ‎2017-04-12

@ChadPruden,

Below is the Kiwisyslog tool link for download. I am able to send SNMP trap:

http://www.kiwisyslog.com/products/kiwi-syslog-server/product-features.aspx

Best Regards,

Shiva

ChadPruden · ‎2017-04-12

Hi Sinhaa,

Thanks for the suggestion, but I still have not found success. (see below)

-bash-4.2$ sudo /usr/bin/telegraf --config /etc/telegraf/telegraf.conf -test
* Plugin: inputs.snmp, Collection 1
2017/04/12 11:42:52 E! Error in plugin [inputs.snmp]: agent 10.65.163.71:161: performing get: Request timeout (after 5 retries)
2017/04/12 11:42:52 E! Errors encountered during processing

Are there any pre-reqs on the WFA host (Win2012)?

* I've opened firewall ports as described in the Admin Guide (WFA host 161/UDP/outgoing, SNMP host 162/UDP/incoming). Also opened 162 outgoing after your reply.

* It's not specified in Admin Guide, so I presume that SNMP-Service should not be enabled as a Feature on the Win2012 server.

What is the community string? "public", "rocommunity", "<null>"

Any tips / tools that can just help me know that the SNMP trap is being sent from the WFA host? If I can confirm that, I'll move to troubleshooting the SNMP listener (telegraf).

Perhaps i am confusing SNMP polling with SNMP traps. WFA sends an SNMP trap message (one time event, broadcast over SNMP) correct?

WFA 4.1RC

sinhaa · ‎2017-04-12

@ChadPruden

What is the community string? "public", "rocommunity", "<null>"

---

WFA send SNMP trap on community public.

...

## SNMP version, values can be 1, 2, or 3
version = 2

....

WFA send SNMP v1 traps. You seem to have your listener configured for v2. Try making it v1 and see.

There also is an issue that SNMP are not sent unless Mail is also configured for WFA. Can you try to configure Mail tab and then see if SNMP traps are received.

If this post resolved your issue, help others by selecting ACCEPT AS SOLUTION or adding a KUDO.

ChadPruden · ‎2017-04-12

Thanks Sihnaa - I've tried those things as well. Still no success.

Email notifications have been working for me for a while now. They are enabled.

Here is is my telegraf config file, excluding the output section. Any other suggestions would be appreciated.

 [[inputs.snmp]]
   agents = [ "10.63.160.179:162" ]
   ## Timeout for each SNMP query.
   timeout = "5s"
   ## Number of retries to attempt within timeout.
   retries = 5
   ## SNMP version, values can be 1, 2, or 3
   version = 1

   ## SNMP community string.
   community = "public"

   ## The GETBULK max-repetitions parameter
   max_repetitions = 10




   ## measurement name
   name = "system"
   [[inputs.snmp.field]]
     name = "wfa"
     oid = "1.3.6.1.4.1.789.1.1.12.0"

# sudo /usr/bin/telegraf --config /etc/telegraf/wfa-snmp-telegraf.conf
2017/04/12 13:44:24 I! Starting Telegraf (version 1.1.2)
2017/04/12 13:44:24 I! Loaded outputs: kafka
2017/04/12 13:44:24 I! Loaded inputs: inputs.snmp
2017/04/12 13:44:24 I! Tags enabled: host=lx1002
2017/04/12 13:44:24 I! Agent Config: Interval:10s, Quiet:false, Hostname:"lx1002", Flush Interval:10s
2017/04/12 13:44:40 E! ERROR: input [inputs.snmp] took longer to collect than collection interval (10s)
2017/04/12 13:44:40 I! Output [kafka] buffer fullness: 0 / 10000 metrics. Total gathered metrics: 0. Total dropped metrics: 0.
2017/04/12 13:44:50 E! ERROR: input [inputs.snmp] took longer to collect than collection interval (10s)
2017/04/12 13:44:50 I! Output [kafka] buffer fullness: 0 / 10000 metrics. Total gathered metrics: 0. Total dropped metrics: 0.
2017/04/12 13:45:00 E! ERROR: input [inputs.snmp] took longer to collect than collection interval (10s)

^ repeated the "Send Test Notification" - also invoked some actual workflows.