Estimating Metric Consumption for Dynatrace SaaS (Network Extension - DPS License)

[AskMe-Solutions]

Hello everyone,

I’m planning to implement Dynatrace SaaS with Extension (DPS License) to monitor the performance and health of network infrastructure in our organization. Our current device inventory includes:

• F5 R5600: 4 devices

• Cisco Catalyst 9200 Switches: 172 devices

• Fortinet 400E: 2 devices

• Fortinet 200F: 2 devices

• VMWare Hosts: 71 hosts

We plan to monitor these devices via SNMP, APIs, or vCenter integration, depending on device compatibility. Metrics will be collected using Extension 2.0 or OneAgent remote plugins.

My questions are:

1. Roughly how much DPS (Davis Data Units) for metric ingest and processing should we expect to consume per year for this setup?

2. Can we customize which metrics to ingest (e.g., only CPU, Memory, Interface Traffic, Device Health, etc.) to reduce consumption?

3. Are there any formulas, tools, or best practices for estimating metric ingestion when using Dynatrace Extensions?

4. Has anyone here implemented a similar monitoring setup? If so, could you share your real-world metric consumption experience?

Any advice, recommendations, or reference architectures would be greatly appreciated.

Thank you

[p_devulapalli · Accepted solution]

@AskMe-Solutions It would be tricky to estimate the license consumption for network devices as it can vary widely based on the metrics ingested and the size/type of network device monitored . If you take an example of a Cisco device , the extension has some details around license calculation which you can use to estimate . 

https://www.dynatrace.com/hub/detail/generic-cisco-router-snmp-extension/

You need to decide on what you would like to monitor on a network device , ie., CPU, Memory , BGP, Interfaces etc., and enable the specific feature sets in the extension to limit the collection of metrics . The extension includes the details of the feature sets that you can enable 

Hope this helps