SNMP monitor doesn't take service down

Description

It's a Synology with 5 drives in a RAID. Drive 5 has an outage (see snmpwalk). The poller should walk the OID tree and if one entry doesn't match "1" it should go offline. But it didn't.

Poller configuration:

 

Snmpwalk on .1.3.6.1.4.1.6574.2.1.1.5:

Karaf Shell polling while drive 5 down:

Debug log:

Acceptance / Success Criteria

None

Lucidchart Diagrams

Activity

Show:

Marcel Fuhrmann August 7, 2018 at 1:57 PM

I've expected the walk output, but it's only the setting value. I guess we can close this one. Thanks for your help.

Alejandro Galue August 7, 2018 at 10:53 AM

In terms of using "${walk}" on the "reason-template", I can't see why it wouldn't work according to this:

https://github.com/OpenNMS/opennms/blob/develop/features/poller/monitors/core/src/main/java/org/opennms/netmgt/poller/monitors/SnmpMonitor.java#L189

Marcel Fuhrmann August 7, 2018 at 4:48 AM

Yes. You are absolutely right. It seems I have misunderstood the match-all parameter.  Sorry for the inconvenience.

 

What a coincidence. Tonight a drive in Synolgy died so I could test it again.

Maybe I found an other issue in this setup. I've tried to add the "${walk}" variable in reason-template. The walk didn't appear in the event. But it should, doesn't it?

 

Alejandro Galue August 6, 2018 at 8:05 PM

Could you elaborate more on what's considered up and down for the Synology-Drive-Physical service ?
To me, it should be up, if ALL the entries (a.k.a. match-all="true") has a value of 1. On the other hand, if AT LEAST one has a different value, it should be considered down. Are my assumptions correct ?

Configuration

Details

Assignee

Reporter

Labels

Components

Fix versions

Affects versions

Priority

PagerDuty

Created August 6, 2018 at 6:24 PM
Updated August 7, 2018 at 1:57 PM
Resolved August 7, 2018 at 1:57 PM