Showing results for 
Show  only  | Search instead for 
Did you mean: 

Hello, I'd like to know how does Dynatrace monitor devices that don't have the chance to install the oneagent, but can impact seriously in the aviability of the services, like routers, switches, APM, etc.


Hello, for such devices it is possible to create custom network device:

This will give you visibility about what particular device is part of transaction (normally you see requests to unmonitored hosts).

If your device is able to push some metrics via payload (or you have extra tool that has such ability) you can also configure custom metrics for this device and push them to Dynatrace. This will give you option to monitor status of it. But you will not see logs, transactions on services with code level etc.


That's could mean that if I want Dynatrace to listen to SNMP Traps for
example, It will recognice them and correlate the problems

If you will create custom network device with visible traffic to it from services, Dynatrace will rise problems when those transactions will fail because of ot. I don't know what about correlation of metrics pushed to Custom Network Device. But you should be able to create custom alerts based on those metrics. I don't know how it's relating to SNMP Traps because I've never worked with them.


Here is another question about this topic:


Thank you so much Sebastian. You know, there is a lot of old stuff that you have to keep an eye on over there.

Dynatrace Pro
Dynatrace Pro


Option number one is to use the API:

Option number two is to use ActiveGate plugins:


That's could mean that if I want Dynatrace to listen to SNMP Traps for
example, It will recognice them and correlate the problems

We use ActiveGate plugins to monitor things like f5, DataPower, Juniper and NetScaler where we cannot install an agent.

If you want to listen to SNMP traps you'd need to write your own application that does that and then pushes it to our API. It would only correlate events/metrics from there with problems if you also pass in the IP/Port where the data is coming from and if we see traffic going to that IP/Port (so it won't work for switches and routers, but it will for non-transparent firewalls, load balancers and proxies.

Thank you so much Michael.

As you may know there is a lot of old stuff over there that only is
able to send SNMP Traps (even not manage snmpget calls!!!) and I think
thats could be very useful if you can use these traps to help in knowing
the root cause of a problem and the services/users impacted by it.

As far as I know there's a few open source developments that can get
the traps and translate in sort of way using mib files. Could it work

I'm running into Business Service Management (BSM) that have something like I'm looking for, but there's not much information over there. Is it still supported?


BSM is no longer supported since a couple of months. A lot of use cases are already covered by Dynatrace in a much better fashion.

Depending on the sources of the SNMP traps it might be a good idea to get them into Dynatrace, or not. We prefer to focus on data that will be consumed by the AI, in which case the data source needs to provide the necessary context for our model to be able to link it to the existing processes and services.