Host Monitoring modes overview (DPS)

  • Latest Dynatrace
  • Overview
  • 5-min read

Dynatrace OneAgent offers four monitoring modes that you can choose from. Some modes offer complete insights into your observability data, while others collect more limited information. For each OneAgent installed in your infrastructure, you can choose the monitoring mode that best fits your needs.

This page describes the different monitoring modes and the features that they provide with a DPS subscription.

For information about how usage of a monitoring mode translates to consumption of your DPS license commit, see

Full-Stack Monitoring feature overview

Full-Stack Monitoring mode offers comprehensive application performance monitoring, including distributed tracing, code-level visibility, CPU profiling, memory profiling, and deep process monitoring for hosts and containers.

  • Host-based Full-Stack Monitoring offers

  • Container-based application-only Full-Stack Monitoring offers comprehensive application performance monitoring.

Metrics

Full-Stack Monitoring lets you use the following metrics:

  • Infrastructure Monitoring metrics
  • Application performance monitoring metrics
  • Other built-in metrics
  • Custom metrics

Distributed traces

Full-Stack Monitoring trace data comes from two sources and takes one of two forms.

  • Trace data from the Dynatrace OneAgent: OneAgent automatically manages the volume of captured trace data via Adaptive Traffic Management. It automatically and continuously adjusts the sampling rate in an intelligent way and keeps the ingested trace data volume roughly within your included trace-data volume.

  • The second data source is trace data sent via the OneAgent Trace API or, more generally, OpenTelemetry traces that originate at a Full-Stack Monitoring host or application and are sent via the OTLP API. Trace data sent this way may be sampled by an OpenTelemetry agent, the SDK, or Collector at a fixed rate. These mechanisms are not controlled by Dynatrace or by Adaptive Traffic Management. This means this trace data is not automatically kept below the included trace data volume.

Dynatrace retains all traces ingested from your environment for 10 days.

Dynatrace provides the ability to extend trace retention on a selective basis for up to 10 years. This is achieved by creating custom buckets in Grail. To request Extended trace ingest for Full-Stack Monitoring, Please contact a Dynatrace product expert via live chat within your environment.

CPU, memory, and thread profiling

Full-Stack Monitoring includes CPU, memory, and thread profiling for technologies like Java, .NET, Go, Node.js, and PHP. OneAgent uses an intelligent patented mechanism to manage the volume of profiling data.

Infrastructure Monitoring feature overview

Infrastructure Monitoring mode provides comprehensive host monitoring for physical and virtual hosts.

In addition to all features of Foundation & Discovery, Infrastructure Monitoring also includes detailed process performance metrics, disk performance metrics, process-to-process network analysis, and per process memory analysis. Dynatrace Extensions can be enabled on hosts with Infrastructure Monitoring mode.

Foundation & Discovery feature overview

Dynatrace OneAgent can be configured for Foundation & Discovery mode, which provides basic monitoring for your hosts (for example, host health, disk status, and OS service status). Unlike other tools that provide basic monitoring, Foundation & Discovery leverages core OneAgent features: discovery and topology.

Foundation & Discovery mode detects process-to-process communication and populates the Smartscape topology accordingly. This provides important clues for AIOps which is included with OneAgent, see Davis® AI automatic root cause analysis for details.

Broad deployment of Foundation & Discovery mode enables you to select the right monitoring mode for each of your hosts. A host’s criticality can be determined based on processes, technologies, externally accessible services, and topological connections.

Mainframe Monitoring feature overview

Dynatrace Mainframe Monitoring provides automatic end-to-end application performance monitoring for transactions, regions, and apps deployed on IBM z/OS. It includes distributed tracing, metrics, topology, and code-level insight for 30+ supported technologies.

With the DPS capability for Mainframe Monitoring you get flexibility in licensing for incremental monitoring rollouts (for example, you can start small and expand coverage over time).

The technical prerequisites for DPS are:

A monitored Logical Partition (LPAR) is represented as a host in Dynatrace. The billing for monitoring an LPAR depends on the partition’s Million Service Unit (MSU) value and the duration of Dynatrace monitoring. An MSU is an IBM measurement of the amount of processing work an IBM Z mainframe can perform in one hour.

Metrics

Mainframe Monitoring includes application performance monitoring and related built-in metrics.

Distributed traces

Dynatrace retains the total amount of ingested trace volume from your environment for 10 days. Dynatrace provides the ability to extend trace retention on a selective basis for up to 10 years. This is achieved by creating custom buckets in Grail.

Related tags
Dynatrace PlatformInfrastructure Observability