Dynatrace SaaS is the default and most common deployment, but Dynatrace Managed offers a compelling alternative when you need more control over your environment.
Let’s see Dynatrace in action. Imagine you’re monitoring a complex microservices application.
graph LR
A[Frontend Service] --> B(Auth Service)
B --> C{Product Catalog}
C --> D[Order Service]
D --> E[Payment Gateway]
E --> F[Database]
When a transaction flows through this, Dynatrace captures every hop, every millisecond spent, and every error.
Here’s a peek at what that looks like in the UI (simplified):
Service Health: Auth Service
- Availability: 99.98%
- Performance: 250ms (avg response time)
- Errors: 0.01% (1 error in last 5000 requests)
- Top Transactions:
POST /login(40% of traffic),GET /users/{id}(30%) - Dependencies:
Frontend Service(incoming),Product Catalog(outgoing)
Dynatrace SaaS is a fully managed service. Dynatrace operates and maintains the infrastructure, the platform, and all its components. You get access to the latest features and updates automatically. This means your team can focus solely on analyzing the data and driving business value, without worrying about patching servers, managing storage, or scaling the Dynatrace cluster itself. It’s like having a dedicated Ops team for your monitoring platform, handled by Dynatrace.
Dynatrace Managed, on the other hand, is a self-hosted solution. You deploy and manage the Dynatrace cluster on your own infrastructure, whether that’s on-premises or in your own cloud environment (AWS, Azure, GCP, etc.). This gives you complete control over the underlying hardware, networking, and operating system. You decide when to apply updates and have direct access to the cluster for deep customization and troubleshooting. This is ideal for organizations with strict data residency requirements, specific security policies, or a need to integrate Dynatrace deeply into their existing infrastructure management practices.
The core problem Dynatrace solves is the complexity of modern, distributed applications. In a world of microservices, containers, and cloud-native architectures, understanding performance and user experience across hundreds or thousands of components is a monumental task. Dynatrace automatically discovers all these components, their dependencies, and the transactions flowing between them. It then uses AI (Davis) to pinpoint the root cause of any issue, whether it’s a slow database query, a misconfigured load balancer, or a buggy piece of code in a specific service.
For Dynatrace SaaS, the primary levers you control are your licensing, user permissions, and the configuration of what you want to monitor (e.g., enabling specific OneAgent features, setting up alerting rules, defining custom user sessions). You don’t touch the infrastructure.
With Dynatrace Managed, you control everything. This includes:
- Infrastructure Sizing: Choosing the right number and type of nodes for your OneAgent ingest, processing, and storage needs.
- Network Configuration: Ensuring proper connectivity between your monitored hosts and the Dynatrace cluster, and between the Dynatrace cluster nodes themselves.
- Storage Management: Allocating and monitoring disk space for metrics, traces, and logs.
- Update Schedule: Deciding when to upgrade your Dynatrace cluster to new versions.
- Security Policies: Implementing your organization’s specific security hardening and access control on the Dynatrace infrastructure.
The difference in how data is ingested and processed between SaaS and Managed is largely transparent to the end-user analyzing data. In both cases, OneAgents send data to ingest nodes, which then process it into metrics, traces, and events. The primary distinction is who manages the availability, scalability, and maintenance of those ingest nodes and the subsequent processing and storage infrastructure. For Managed, your team is responsible for the health of the Dynatrace cluster itself, while for SaaS, Dynatrace is.
When configuring Dynatrace Managed, pay close attention to the network latency between your OneAgent-instrumented hosts and the Dynatrace cluster. High latency here will directly impact ingest performance and can lead to data gaps or delayed availability of metrics, even if the cluster itself is healthy. You can monitor this through the Dynatrace cluster’s own dashboard, looking for metrics like network.receive.latency and network.send.latency on the ingest nodes.
The next step after choosing between SaaS and Managed is understanding how to effectively license and scale your OneAgent deployment to match your monitored environment.