Sizing calculator for Dynatrace Managed

[ashutosh_singh1]

Is there any sizing calculator available for Dynatrace Managed?

[skrystosik · Accepted solution]

Hello, there is no calculator, but here is really useful table:

https://www.dynatrace.com/support/help/deploy-dyna...

It is hard to calculate such things with 100% accuracy because of many factors that are responsible for this. Generally if I can advice something, focus on hardware requirements, storage is easiest to extend later. You can also tweak data retention time in Cluster Management Console to use as much of your storage as you can 🙂

Sebastian

[ashutosh_singh1]

Hi Sebastian,

Thanks for your response..!
I have gone through with hardware sizing. My concern is here, Suppose i have 150 agents so i ll go with medium node type.
If i have 150 agents with 2 lacs user action/Minute then which node type will be applicable?

Kindly let me know your thoughts.

Regards,
Ashutosh Kumar Singh

[skrystosik]

150 hosts monitored by oneagent should be covered by small instance, but If you are planning extending number of agents in the future, you can go with medium. Generally we are using this table only as a starting point. We are also thinking about future. In our clients environments it is hard to change resources of cluster (sometimes), so we are trying to look into the future. But if your cluster will be VM with option to add cores and ram dynamically I think you shouldn't be afraid 🙂

Sebastian

[ashutosh_singh1]

Hi Sebastian,

I agreed with your point but my concern is here for 200000 user action/min. Does it affect resource utilization. If yes, so how to calculate sizing for above mentioned user action with medium.

Thanks,
Ashutosh

[skrystosik]

Ok but what does it mean 2 lacs action/min because this is what I don't understand 🙂 Generaly amount of user actions per minutes from those hosts is consuming resurces because all transactions recorded by Dynatrace has to be collected processed and stored. User action is page load / click on some elements etc. If you don't have such data I recommend starting with some configuration base on hosts number and check if it's performing well.

Sebastian

[ashutosh_singh1]

Same i was trying to say here, I have an 150 agents and its user action(user clicks) is more than 200000 lakhs/minute.But as mentioned in hardware sizing documentation,medium node can handle peak user actions/min (per node) is only 25000.

In my scenario agents count are according to medium node capacity but peak user action far more as shown in sizing table. In this case how we have to go ahead with sizing.

I have to hit and try or there are few parameters on that we can calculate.

Thanks,
Ashutosh Kumar Singh

[skrystosik]

If you have such huge amount of user action on your enviornment I think it will be better to consider multi cluster environment. But I think, about sizing such configuration it will be better to talk with somebody from dynatrace.

Sebastian

[mevlut_soner]

Hi,

There is a size calculator tool. For more specific requirements, the following information is required. For this, contact Dynatrace Support Team and share following informations.

• Number of hosts
• Avarage number of monitored processes
• Number of nodes in cluster
• Clockspeed per CPU
• Service calls per minute
• User actions per minute
• Logs storage per day (GB)
• The amount of replayable sessions per day
• Retention times for all of this