1.1 Dynatrace Platform Capabilities, Web UI Navigation & the Dynatrace Hub
Key Takeaways
- Dynatrace groups its value into four capability pillars — Observability, Security, Automation, and Integration — and the Associate exam's largest scored topic, Foundational Platform Capabilities, tests your ability to name them and place a feature under the right one.
- The modern platform is app-based: use-case apps such as Infrastructure & Operations, Kubernetes, Distributed Tracing, Logs, Vulnerabilities, and Business Flow sit on top of five core technologies — Grail, DQL, Davis AI, AppEngine, and AutomationEngine.
- Grail is the schema-on-read data lakehouse that stores logs, events, business events, traces, and metrics in one place, and DQL is the single query language used to read all of it.
- The Dynatrace Hub is the in-product marketplace for apps, extensions, and technology integrations; it is where you activate a capability rather than installing separate software.
- Dynatrace Documentation, the Support Center, the Community, Dynatrace University, and the free Playground tenant are the five official resources the certification learning path expects you to be able to name.
Before you can deploy a OneAgent or write a single line of DQL, the Dynatrace Associate exam expects you to be able to answer a deceptively simple question: what is Dynatrace, and where do I click? Foundational Platform Capabilities is one of the twelve scored topics on the exam score report, and it is consistently one of the heaviest. Candidates who jump straight into deployment mechanics without learning the platform's vocabulary lose easy points on questions that ask which capability solves a given business problem, or which app a piece of data lives in.
The Four Capability Pillars
Dynatrace describes itself as a platform that delivers analytics and automation for unified observability and security. Its own learning path breaks the value proposition into four pillars, and exam questions frequently ask you to classify a feature into one of them.
| Pillar | What it means | Representative features |
|---|---|---|
| Observability | Real-time monitoring and visualization of applications, services, and infrastructure | OneAgent, PurePath distributed tracing, Smartscape, Real User Monitoring, Synthetic |
| Security | Runtime vulnerability detection, attack protection, data privacy, and compliance | Runtime Vulnerability Analytics, Application Protection, Security Investigator |
| Automation | AI-driven discovery, analysis, and no-code/low-code remediation | Davis AI, AutomationEngine, Workflows, auto-instrumentation |
| Integration | Ingesting and correlating data from clouds, containers, and open standards | Cloud integrations, OpenTelemetry, Extensions, OpenPipeline |
The unifying claim is that these are not four products bolted together — they read from the same topology and the same data store, which is why Dynatrace can tell you that a vulnerable library is running in a process that serves a revenue-critical checkout service.
Core Platform Technologies
Underneath the apps sit five named technologies. Memorize what each one is, because distractor answers routinely swap them.
- Grail — the schema-on-read data lakehouse. Logs, events, business events, spans, and metrics all land here. Grail requires no up-front schema definition, unlike a relational database.
- DQL (Dynatrace Query Language) — the single query language used to read everything in Grail. One language for logs, traces, metrics, and business events.
- Davis AI — the causation engine that turns raw anomalies into a small number of ranked problems with an identified root cause.
- AppEngine — the framework that lets customers and partners build and share custom apps on top of platform data.
- AutomationEngine — the no-code/low-code engine behind Workflows, used to act on observability and security insights.
Two more names appear constantly: OpenPipeline, which processes, enriches, filters, and masks data on ingest before it is stored, and Smartscape, the automatically discovered topology model that gives every entity its relationships.
The App Model and the Web UI
The modern Dynatrace UI is a collection of apps rather than a fixed menu tree. From the Launcher you open the app that matches your use case:
- Infrastructure & Operations — host, process, disk, and network health
- Kubernetes — clusters, nodes, namespaces, workloads, pods
- Distributed Tracing and Services — request-level analysis
- Logs and Notebooks — investigation surfaces
- Dashboards — continuous monitoring surfaces
- Problems — Davis-detected problems and root causes
- Vulnerabilities and Attacks — Application Security
- Business Flow and Business Analytics — business events
- Settings — configuration, and Account Management for licensing
A point of real confusion — and a favourite exam distractor — is that some screens are labelled Classic. Dynatrace is mid-migration from the older, page-based experience (Hosts Classic, Metrics Classic, Synthetic Classic, RUM Classic) to Grail-backed apps. When a question contrasts "the Hosts page" with "the Infrastructure & Operations app," it is usually testing whether you know that the newer app reads from Grail and is queryable with DQL.
Global search and Davis CoPilot (natural-language assistance, including DQL generation) are the two fastest navigation aids in the product.
The Dynatrace Hub
The Dynatrace Hub is the in-product marketplace. It is where you browse, install, and manage:
- Apps — first-party and partner-built UI experiences
- Extensions — technology-specific data collectors (databases, message queues, network devices, VMware, SNMP)
- Technologies and integrations — cloud services, CI/CD tools, ITSM connectors
The exam framing to remember: you do not download and install separate agents to add a capability. You activate it from the Hub, and OneAgent or an ActiveGate picks up the new collection job. A question that offers "download the installer from the support portal" as an option for adding, say, MySQL monitoring is nearly always wrong — the Hub is the answer.
Official Resources You Are Expected to Name
The certification learning path explicitly asks candidates to visit and bookmark these:
| Resource | Use it for |
|---|---|
| Dynatrace Documentation (docs.dynatrace.com) | Authoritative product behaviour, limits, and configuration reference |
| Dynatrace Support Center | Support tickets, on-demand courses, live training, certification |
| Dynatrace Community | Peer discussion, tips, product news |
| Dynatrace University | Learning plans, the Associate learning path, exam scheduling |
| Dynatrace Playground | A live, pre-populated demo tenant for hands-on practice at no cost |
The Playground matters practically: the Associate exam includes 10 to 15 practical questions, and the practical section is open book. Candidates who have never clicked through a real tenant lose those points regardless of how much theory they memorized. Every "Call to Action" in the official learning path is a Playground exercise — treat them as exam prep, not optional enrichment.
Exam Traps in This Section
- Grail is not a database you define schemas for. Answers describing table creation or column typing before ingest are wrong.
- DQL is one language for all data types. Distractors that pair a separate query language with logs versus metrics are wrong.
- Davis is not "machine learning that correlates." It is deterministic causation built on the Smartscape dependency graph — covered in depth in Chapter 7.
- The Hub activates capabilities; the Support Center is for tickets and training, not feature activation.
A platform team wants to add monitoring for their on-premises MySQL estate. A junior engineer proposes opening a support ticket to obtain a separate MySQL monitoring agent installer. What is the correct Dynatrace approach?
An architect is explaining why Dynatrace can store logs, traces, business events, and metrics without the team predefining table structures for each data source. Which platform technology is responsible for this behaviour?
During the practical portion of the Associate exam, a candidate must demonstrate navigating from a detected problem to the underlying host metrics. Which combination of platform elements makes this drill-down possible?