A quick reference for basic Dynatrace concepts and terms.
Apdex is a performance-measurement standard that strives to show the relationship between recorded performance measurements and real-user satisfaction. It offers a uniform means of measuring how well performance meets user expectations. For full details on the Apdex standard, please visit Apdex.org. In the context of Dynatrace, Apdex is used to give you a quick and easy rating that you can use to evaluate the satisfaction of your application’s end users. Apdex ratings in Dynatrace are based on the application-specific thresholds. An apdex measurement rating of
1 equates to perfect performance. An Apdex rating below
0.5 equates to poor performance. For more details about how Apdex is applied in Dynatrace, please see What is an Apdex rating?
Applications consist of web pages that are served by web servers and web containers, for example Tomcat. Dynatrace monitors all the individual components that applications are built upon: web requests, database requests, processes, services, and more. It’s these individual components working together that collectively deliver what your end users view as a complete application.
When thinking about monitoring your application it’s helpful to remember that Dynatrace monitors all of your application’s services, processes, and infrastructure. By evaluating all such components collectively, Dynatrace is able to pinpoint exactly how each service contributes to the performance of your application.
Availability is the success rate at a given instant or time period that indicates if your application is fully functional and available to users.
Availability web checks
An availability web check is the equivalent of a simulated user coming from a real, up to date browser. They can be configured to run from any of our global locations at a frequency of up to every 5 minutes. Web checks will let you know when your application is inaccessible from the internet and when your baselined performance is degrading.
Candidates (aka “monitoring candidates”)
Monitoring candidates are hosts (either VMware virtual machines or EC2 instances) that communicate with monitored hosts in your environment, but don’t themselves have Dynatrace OneAgent installed. It’s recommended that you install Dynatrace OneAgent on all monitoring candidates to gain full visibility and complete monitoring capabilities.
In Smartscape, candidates are visualized with dashed circles and a generic host icon. They are shown linked to the relevant datacenter.
Inactive monitoring candidates (those that haven’t communicated with a host for more than 2 hours) are not included in Smartscape or listed on the Hosts page.
Clickpath web check
Clickpath web checks are simulated user visits that monitor your application’s business critical workflows. Use the Dynatrace recorder to record an exact sequence of clicks and user input that you are interested in monitoring for availability and performance. Once you’ve captured the mouse clicks and additional user actions that you want your clickpath web check to include, your web check will run automatically at regular intervals to test your site’s availability and functionality.
In Dynatrace Smartscape technology, a datacenter can be either:
- A grouping of virtual machines running in an Amazon cloud instance. We map Availability Zones to your data center.
- Or a set of vCenter-based virtual hosts that transmit their data to Dynatrace via a single Security Gateway. You may have multiple vCenter servers or standalone ESXi hosts associated with a single datacenter.
Each virtual host within a datacenter must have Dynatrace OneAgent installed on it. You will see at least one datacenter in Smartscape view if you have a Security Gateway installed. If you don’t have a Security Gateway in your environment we won’t be able to create a datacenter for you.
Note that we provide real-world geographic location details for data centers—look for datacenter location details in Smartscape and on Host pages (in the Properties pane).
Duration is the amount of time a web check or synthetic action waits before it is a possible to interact with the web application.
Deep process monitoring
Dynatrace provides you with deep visibility into the performance of all the services that comprise your web application: web services, web containers, database requests, custom services, and more. You can even track service performance over time, with both historical and up-to-the-moment data.
An “event” reported by Dynatrace is a manual action performed in your environment. Examples include a machine reboot, system shutdown, process restart, or new code deployment. Any manual action you perform on a server is an event, even if it’s a regularly scheduled event.
Any service that is called by your application or another service, but isn’t directly monitored byDynatrace OneAgent is considered to be an external service. Dynatrace maintains awareness that external services are in operation because these services are in communication with the services and/or applications that Dynatrace does monitor, however Dynatrace doesn’t receive performance data related to these services. An exception to this rule is database services. If you have Dynatrace OneAgent installed on a machine from which JDBC calls are sent to your database,Dynatrace will provide a full spectrum of performance analysis related to the called database service (even when Dynatrace OneAgent isn’t installed on the database server machine itself).
There are a number of variables or “dimensions” that affect application performance—most notably, geographic region, OS, browser, connection type, and user action. These are the five dimensions that Dynatrace uses to evaluate application health. Each subset of user is subject to a unique set of these variables so Dynatrace measures the health of and reports on each of these user subsets separately.
A “problem” in Dynatrace includes the AI-driven analysis, environmental context, root cause analysis, and other details provided for one or more incidents in your environment. Problems can express themselves in your environment as performance degradations, improper functionality, or lack of availability. A problem can be the result of a single “event” or multiple events.
A process is an instance of a computer program that is being executed. Processes serve as containers that host services. When you look at processes you see topology information, whereas services give you code-level insight. For example, you might have a Tomcat process that hosts a web application in the form of a server-side service. While processes are host-centric—associated with a single machine in your environment—services are request-centric and therefore typically span across multiple machines in a data center.
A “process group” is a set of processes that perform the same function across multiple hosts. For example, you might have a cluster of servers with each server running the same process in support of multiple hosts.
Real-user monitoring shows your visitors experience when they interact with your application. It is a combination of web analytics and performance measurements that build the picture of how your customer-facing components operate.
The Dynatrace OneAgent monitors the performance of the computers in your application-deployment environment and all their processes. Based on your environment settings, Dynatrace OneAgent enables detailed performance monitoring and real user monitoring of your application and its services. Changes to your applications as well as infrastructure and performance problems are reported to Dynatrace for performance monitoring.
“oneagentmon device” appears in your Windows system during Dynatrace OneAgent installation. It’s used by Dynatrace for deep process monitoring. It works like a monitoring driver and allows Dynatrace OneAgent to add its own library between the operating system and processes it is running.
Dynatrace Security Gateway
Security Gateway is a Dynatrace component that provides secure and reliable communication with cloud-based Dynatrace cluster nodes. Similar to a proxy, Security Gateway reduces bandwidth requirements by encrypting and compressing network communication. It also provides monitoring of virtual infrastructure. Stopping or disabling Security Gateway reduces the monitoring capabilities of this and other hosts that use Security Gateway to communicate with Dynatrace clusters.
A session (i.e., a visit) covers any interaction between a user’s browser and your application. A session includes at least one action. Sessions end when a user’s browser closes or is inactive for 30 minutes. Note that bounced sessions (sessions with only a single action) don’t count against your billable session quota.
Services (aka “Server-side services”)
Web applications consist of web pages that are served by web servers and web containers, for example Tomcat. The web requests that are sent to a specific Tomcat server are an example of a server-side service. Web and mobile applications are built upon services that process requests like web requests, web service calls, and messaging. Such “server-side services” can take the form of web services, web containers, database requests, custom services, and more. Services may in turn call other services such as web services, remote services, and databases services.
Processes are essentially containers that host services. When you look at processes you see topology information, whereas services give you code-level insight. For example, you might have a Tomcat process that hosts a web application in the form of a server-side service. While processes are host-centric—associated with a single machine in your environment—services are request-centric and therefore typically span across multiple machines in a data center.
Service backtrace visualizes all the services that send requests to a specific back-end service. More than just showing which services directly call a specific back-end service, Service backtrace shows you the sequence of service calls that leads up to each request, all the way back up to the browser click that triggered the sequence of calls.
Using proprietary Smartscape technology, Dynatrace is able to automatically discover your entire application stack, from the datacenters, hosts, and processes that comprise your infrastructure and network, through the services that comprise your applications, and extending all the way up to your applications. Dynatrace even discovers and maps all of the dependencies of these components, in real time. Dynatrace then presents its findings to you in an interactive Smartscape topology map. You can click anywhere within a Smartscape visualization to view performance statistics for each of discovered entity in your environment.
Synthetic actions are the equivalent of simulated user actions. When a clickpath web check triggers the loading of a new page or an XHR request, we monitor the availability and performance of that action.
User actions are the actions your customers perform within your application. User actions equate to common user activities such as performing a search, viewing an account balance, viewing items in a shopping cart, or ordering a product. User actions vary based on the features of your application.
The duration of a user action is called action duration. This represents the time that the user must wait before they can proceed. So a low action duration is better than a high action duration. For more details, see What are user actions?
“Wall clock time” is the measurement of time beginning when a computer-based task is started and extending all the way through until the task is completed. Wall clock time is what humans experience as the passage of time and measure using a chronometer, for example a clock mounted on a wall. Wall clock time is the time as perceived by human beings, minutes, seconds or hours. Measuring application performance using wall clock time is useless mainly because of the application components running in parallel.