You are here

Interface passive SNMP check and poller

6 posts / 0 new
Last post's picture
Interface passive SNMP check and poller

Hello forum,

I'm migrating hosts from my old community Opsview to this brand-new MSP Opsview.

I'm a little confused about Interface SNMP check management.

The previous one was simply an active check, polling for the interfaces previously quesried. This new one, if I understood, takes all the data from interface poller, with just 1 quesry for all interfaces, and distributes these data to the passive "interface".

Now... I get an "unknown" alarm for that IF that doesn't have any throughput, where, instead, in the previous check they displayed a more normal 0/0. By now I bypassed this alarm creating a new check, interface, similar to the other one, active and so on, but if so, this check makes a poll for every interface.

I'm missing the concept - could I use this new checking way, but not receiveng an "unknown" if an IF has no throughput (maybe it's just for a few seconds, but it's working)?

Thanks a lot



matthew.kelley's picture
Re: Interface passive SNMP check and poller

I would first go to view/edit host interfaces and query the device.  Make sure that your checks are looking at interfaces that have data available for them.

If the more important thing that you want to achieve is restrict notifications for temporary events, you can introduce a buffer between when any event (warning, critical, unknown) and when you are alerted about it.  This is accomplished by increasing the max_check_attempts field in the service check.

I would also suggest that you raise a formal support ticket on the issue if this needs further attention.


Matt's picture
Re: Interface passive SNMP check and poller

Hi Matt,

no prob for the ticket, but I think this is not an issue. I mean, I'm quite sure this is smthg that I didn't understand.

Community - I used the check "Interface": it was an active check, queried SNMP every IF I decided from the "query" page. And it 1) raised an alert if the IF was down; 2) made a graph of traffic.

But if the IF was NOT down, and simply no traffic for whatever reason, I didn't recevie an alert. Could I do the same with the new couple of checks, IF poller + Interace (passive)? No prob if throughput 0, but no "unknown" if this happen, unless the IF is down.

Thank you for your update Matt.

Raffaello's picture
Re: Interface passive SNMP check and poller


as supposed, it was my understanding. Opening Check help I found the option "--ignore-no-throughput" to add to the check to avoid that "unknown" status in a no throughput case.

Of course I would open a ticket in case of issues. But as told in my previous post, it wasn't a issue. Should we open a ticket for any clarification? What's the purpose of this forum so?

Thank you



smarsh's picture
Re: Interface passive SNMP check and poller

Hi Rafaello,

This forum is for those who want to share ideas / suggestions, or for Core users who want help from other users as they are not paying for Support.


If you raise problems here, the likelyhood is that these issues wont be fixed for you quickly - therefore it is very important you raise support tickets and take advantage of one of the main reasons you have an Enterprise subscription.

All the best,

Sam's picture
Re: Interface passive SNMP check and poller

Dear Sam,

I've been a forum and opsview User since beginning of 2010, I had the "ramboy" account name. I think to know the purpose of this forum, even because more than once I helped (and was helped too of course) in problem solving. For sure I helped with little 2 cents to develope the commercial version.

My original post was a simple question, not raising a problem, a kind of "how this works?". I never wrote "urgent", or "hurry hurry". Just to understand. Of course in case of blocking issues I won't hesitate to raise a ticket. As you told, this forum is an open source way to share ideas and suggestion and, why not, solve "not urgent" problems. This was my case. If you read my post (but for sure will be me being not clear) I found a workaround, in the while I found explaination by myself.

Many thanks for your feedback, my best regards,