1. The old license model (Separate components) #
Historically, Dynatrace offered a model where licenses were split into several specific quotas (Units).
Segmentation by Units #
In this model, you had to buy and allocate a precise volume for each type of usage:
- Host Units: Used to monitor infrastructure and applications. Each host is assigned a number of units based on the amount of RAM (e.g., a 16 GB RAM host = 1 host unit).
- DEM (Digital Experience Monitoring): Used for Real User Monitoring (RUM) and Synthetic Monitoring (Robots).
- DDU (Davis Data Units): Used for collecting custom metrics, logs, traces, and events.
- AppSec (Application Security): Units allocated to detecting and analyzing application vulnerabilities.
Assessment of the old model #
Advantages: Strict control and predictability by silo. You know exactly how many servers or log volume you are allowed to monitor, which limits unexpected financial overruns for a specific feature.
Disadvantages: High rigidity (silo effect). If you consume all your DDUs for log ingestion but have unused 'Host Units' left, you cannot transfer this budget from one pillar to another. This slows down the adoption of new features and causes heavy administrative management (regular contract amendments).
2. The new model: DPS (Dynatrace Platform Subscription) #
To adapt to Cloud environments and unify billing, Dynatrace introduced the DPS model.
The concept of global commitment #
The DPS model puts an end to siloed management. The company commits to a global annual budget (a monetary "Pool"). All platform features (host monitoring, traces, logs, security, DEM) are valued according to a pricing grid (Rate Card) and dynamically consume this shared budget on demand.
Assessment of the DPS model #
Advantages: Absolute flexibility. The company spends its budget where it needs it, when it needs it. This model offers great financial transparency and gives immediate access to all new Dynatrace features without requiring the purchase of new specific licenses.
Disadvantages: Requires strong and proactive governance (FinOps approach). Since all data draws from the same budget, a poorly optimized application that suddenly generates an excessive volume of logs or traces can drain the company's overall budget envelope if no quota or consumption alert is put in place.
