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

Cluster node capabilities

DynaMight Guru
DynaMight Guru

Dynatrace has an option for Cluster node capabilities,

where for instance certain nodes can be dedicated to UI traffic, or to processing OneAgent data.

My main idea is maximizing cluster node usage, while also bringing response times down.

Has anyone used this and have some practical advice on how this would really help? I imagine "it depends", but some factors come to mind:

  • The ratio between processing needed for UI and OneAgents differs from Organization to Organization. Are there metrics that we could rely on to make this configuration?
  • If for instance a node doesn't serve OneAgents, do the required modules for processing OneAgent traffic get unloaded and stop using resources, like memory, for instance?
  • Faster UI response times is one of the announced improvements, but is it better in the case where we disable OneAgent traffic on certain nodes, or we disable Web UI on certain nodes?
  • I imagine that in the typical 3 node cluster configuration, we could not disable OneAgent traffic in one of the nodes, due to the split-brain sync problems?
  • If we would dedicate only one node for UI, I believe that in the case of update for that node, or for some reason it goes down, that the UI would be inaccessible?
Antonio Sousa

Dynatrace Champion
Dynatrace Champion



This is quite interesting, is any kind of DNS load balancing happens for oneagent and UI client traffic on nodes? or does this request gets segregated using any kind or pre-processor on nodes?


DynaMight Guru
DynaMight Guru

Hello @AntonioSousa & @techean 

Did anyone of you try?

According to the below statement, Dynatrace is disabling the Web UI traffic. My idea was to include more nodes with fewer resources to enable only Web UI traffic. 

To improve DNS load-balancing performance, starting with Dynatrace Managed 1.214, web UI traffic is by default disabled when installing the 13th and subsequent nodes (or 7th and subsequent nodes in a data center for Premium HA).

The questions remain the same related to the hardware requirements.




Nodes in a Dynatrace cluster should all be equal in terms of resources:


Regarding the other questions I made, I have not got much feedback.

Antonio Sousa

Hello @AntonioSousa 

Thank you for your comments.

It is the documented thing. I am interested if someone has tried it personally. For example, why do I need to have the same hardware in case I just want to use it for the Web UI traffic when the number of end users increases to use the Dynatrace?



Featured Posts