Minimum Resource Requirements for Dynatrace Managed POC with 10 HU

[agylpradipta]

Hello everyone,

I would like to ask about the system requirements for Dynatrace Managed.

Refer to the following documentation: https://docs.dynatrace.com/managed/managed-cluster/installation/dynatrace-managed-hardware-and-syste...

It states that for a Micro (max 50 HU) setup, the required resources are:

• 4 vCPU
• 32GB RAM
• At least 200GB storage

Currently, I want to do a POC for a prospect, who based on our calculations, will only consume 10 HU.

Since the POC will be conducted in their Dev/QA environment first.

Is it possible to reduce the resource requirements I mentioned above?

If so, what would be the minimum requirements?

Thank you.

[PetrVlacil · Accepted solution]

It is little bit late answer, but If you want to showcase the Log Management in your PoC, you would need a 64GB RAM as the requirement for Log Management Classic is 64GB RAM.

[Peter_Youssef]

Hello @agylpradipta 

Regarding the POC raised concerns:

1. You should consider the technology stack and related integrations you are conducting the POC for showing a practical case.
2. So, based on the customer inputs, we start the planning and sizing accordingly.
3. In case of having multiple cloud technologies: we consider sufficient sizing of DDUs while maintaining the HU and DEM units.
4. In case of RUM and Server insights: consider sufficient sizing of HU && DEM units while considering a small license of DDU for log and integrations if exist.
5. As per the provided input Micro sizing would be sufficient as long as you won't extend to log monitoring as it requires High computational resources "CPU, Memory, Storage".
6. If the POC requirements to test DEV or UAT apps, you can utilize One small sized node instead of having three 3 micro sized nodes.
7. It's recommended having three cluster nodes of the same size either micro or small.
8. You can utilize these nodes after the POC in the Production, what you will ask for is adding extra resources equally to reach Medium or large node size according to monitoring requirements.
9. So you can make best use of the POC resources either to uninstall and start from scratch again or perform cluster backup && Restore to install the Production cluster nodes.
10. It's not recommended to decrease the provisioned resources, as long as the monitoring requirements increase as per customer requirements, you will ask for more resources to ensure the platform working as expected. 

So:

• As per the current inputs proceed with the small sizing, no need to decrease the provisioned or suggested resources.

• Extended monitoring requirements, consider Small sizing:

• For Production deployment, it's recommended as per the practical use cases to proceed with at least 3 cluster nodes "M" or "L" based on customer requirements.

Tip:

• Having sufficient resources will make it easy to conduct POC activities smoothly without monitoring issues due to resources lack.
• Delivering customer expectations and going beyond depends on the platform stability, affordability, and resilience against the excessive monitoring requirements, which in turns being translated to Platform computational resources.

Managed requirements:

• managed-requirements 
• log-monitoring 
• managed-system-requirements 

Wishing you successful POC.

KR,

Peter.

[fstekelenburg]

One question that springs to mind, reading this thread, is: is there a specific need to PoC with a Managed set up, would a SaaS driven POC not be an option?

Adding to all the valueable information recorded here, from experience: a cluster is not required for a PoC, one node is sufficient. Also for running a very small environment. Go for a "small" factor node. If in the end you want to make it into a managed cluster (min. 3 nodes), small is the mininum requirement per node. 

How did you turn out with the POC?