dynaTrace software GmbH - dynaTrace software

Skip Menu

dynaTrace is the new generation of application performance management. Our patented PurePath technology provides the world’s most accurate and detailed view of application behavior. All transactions, end-to-end. From user click, across all tiers, to the database, and back again.

With PurePath, we help with the one thing that developers, operators and business owners all agree on: applications performance matters. Meet dynaTrace and experience the new generation of APM.

We regularly get questions on dynaTrace, the PurePath technology, and best practices. Therefore, we have compiled a number of frequently asked questions. Check out the answers below.

Can dynaTrace be used in production?

What overhead should be expected?

Why is it essential to trace and capture all transactions continually and not just on-demand or via sampling?

What is the secret sauce of PurePath Technology?

Is dynaTrace using a sampling approach to scale to large environments?

Does dynaTrace provide automatic entry-points and business transactions to get started quickly?

Does dynaTrace provide automatic baselining?

Do I need a separate UEM tool to have both APM and UEM?

Does dynaTrace tell me the relationship between real user experience and business impact?

How does dynaTrace handle monitoring of large environments beyond 1000 JVMs/CLRs/Web Servers?

Does dynaTrace support highly, globally distributed SOA environments?

Does dynaTrace support cloud environments?

How does dynaTrace support new technologies like Big Data solutions?

Does dynaTrace support the agile development process?

Does dynaTrace provide easy access to data for those who are not experienced with dynaTrace?

Yes, dynaTrace has been proven for years in large and demanding production environments, with ultra-low overhead, capturing its patented PurePath™ data for all transactions, end-to-end, 24x7:
1. 9M transactions an hour on Black Friday for one of the top 5 retailers in the USA
2. Over 1600 production JVMs for the largest Wall Street trade clearing house 24x7x365
3. 2500 JVMs & .NET instances throughout the application lifecycle (production, test, development) for one of Europe’s largest banks
4. "Sentry Insurance Boosts Performance with Application Monitoring Solution" (Aug 22, 2011)
5. Gartner has acknowledged the fast growth and innovation of dynaTrace and recently, after multiple production customer interviews, placed dynaTrace into the top visionaries quadrant, September 2011 (read here)
Also, dynaTrace is the only production APM system that has been purpose-built to support the full performance lifecycle. dynaTrace provides three editions uniquely optimized to users and environments in Development, Test Centers and Production and supports both classic and agile approaches:
1. dynaTrace is the only solution on the market that allows sharing of deep-dive production information such as offline session files across the software application lifecycle for the fastest MTTR on the market and for maximizing collaboration.
2. dynaTrace Development Team edition is used for Automated Performance Regression Analysis in Continuous Integration, Profiling, Architectural Validation, Team Collaboration and more.
3. dynaTrace Test Center edition is most commonly used with its integration into Load Testing tools/environments to reduce the number of test runs and to accelerate diagnostics.
4. dynaTrace Production Edition is used by Business Owners, Operations, Performance Analysts, Marketing, IT management, Services/HelpDesk, and other areas of the business due to its unique capability to bring application performance management together with business needs and visibility.

2% or less. dynaTrace features a unique ultra-light agent technology, that offloads all heavy number-crunching to the dynaTrace Collector and Server, thereby having the lowest memory, response-time and CPU overhead while capturing transactions 24x7, always on.
Zero-configuration and auto-discovery mechanisms ensure hassle-free deployments, auto-adapts to changing code – even code that’s hot-deployed at runtime. dynaTrace automatically manages its overhead such that it remains below 2%. At the same time dynaTrace does not miss a single transaction and it also does not perform simple sampling, a low value approach used by other vendors whose architecture and/or approach preclude production safe, all-transaction use.

Mapping IT to business is more important than ever, to drive success and to retain competitive advantage. For example it is important to understand the impact of user-experience versus revenues. In order to calculate the total revenue, all transactions have to be observed (e.g. calculating a daily total revenue total).
Conversion rates are very important metrics to monitor in order to drive the business. By definition it is the ratio of total transactions to those transactions that convert. Naturally this means that all transactions need to be captured/traced and analyzed to calculate the converted ones.
Enabling complaint resolution which requires both good and bad transactions to prove whether indeed a transaction was bad or simply a false complaint.
Understanding your users' click path through your system. Only with all transactions available is it possible to understand how users got to the point of failure or to the point where they abandon the site.
Capture the root-cause of a distributed transaction that’s an outlier, not a root-cause of transactions that are captured AFTER a threshold breach triggers that more frequent sampling is needed.
Architectural validation and improvement can only be done based on real-world production data. By analyzing trends, real-user usage patterns, performance and scalability implications of your production system, you can validate and improve the architecture of your applications preventing issues proactively, before they cause impact to users.
Measuring the performance of business transactions that do not have the required information at the first entry point (e.g. tenant information that’s on the 3rd tier, e.g. top transactions that call more than 100 SQLs on the backend tier, which user caused "n" CICS transactions).
Understanding scalability and synchronization issues requires the understanding of how transactions impact each other and where the race conditions within these transactions are. Only continuous transaction tracing can deliver the required level of detail to identify both the blocked transaction and area of code, and the transaction and area of code that is responsible for the block.

PurePath provides true distributed tracing across Web/WebServer/Java/.NET/C/CICS; not a sub-trace per JVM/CLR and then stitched together, which in turn is not able to relate a slow end-user response time measured on Tier1 with a poor-performing database access on Tier2, because so called “smart agents” aren’t that smart since they have no clue what other tiers are doing. That is a key reason why the real smarts needs to reside in the collection and server technology.
PurePath also captures contextual information such as User Session Information, Method Arguments, Return Values, Log Messages, Exception Details, etc.--this is important for architects/developers to identify the true difference between a good and bad transaction.
A single PurePath or a set of PurePaths can be exported to a dynaTrace Session and shared with your colleagues or 3rd Party providers. The exported dynaTrace Session not only includes the full transactional trace and all contextual data, it also contains every system metric captured in the environment at the time of the transaction, e.g., CPU, Memory, Application Server Metrics, etc.
Data captured along the PurePath can be used for SLA management (e.g., execution time to external web services) or can be used for Business Transaction Management (e.g., Revenue by User per Geographical Location).

No. Unlike other vendors, who use a sampling technique down to one sample per minute to cope with large environments, dynaTrace ensures that it does not miss any business metric, performance metric, and good or bad transaction. dynaTrace scales with ultra-lightweight agent technology, zero-config. auto-sensor capabilities, and highly tuned,cloud optimized and distributed dynaTrace Collector and Server architecture. Also see "why it is essential transactions are always on" above.

Yes, dynaTrace auto-detects Web, Web-Services, RMI, JMS, JNDI, ASP, Web-Browser clicks, and many more entry points and provides auto-business transactions and auto-dashboards, for example, the list of worst-performing Web pages, Landing pages, Business transactions and more.
dynaTrace customers are not limited to these entry points, as dynaTrace has an excellent auto-discovery and point-click UI to configure proprietary entry points.
In addition, dynaTrace allows easy configuration of business transactions specific to your line of business, for example, “Booking Revenue per Destination and Tenant”. This enables our customers to measure and quantify business performance with application performance.

Yes, for User Experience performance metrics, where a true competitive Speed of the Web baseline is determined daily by industry.
Soon, in spring 2012, dynaTrace will provide a unique and innovative automatic threshold establishment for server-side metrics that goes far beyond the classic time-slotting of historical metrics for predicting the future (The classic time-slotting approach is like trying to predict the stock market today purely based on historical data.)

dynaTrace has extended its PurePath Technology into the Browser thereby delivering the industry’s first fully converged server-side APM and client-side UEM solution.
There is no separate tool, installation or data repository. With dynaTrace you use one salutation that combines and links UEM with APM allowing you to analyze the impact of server-side performance and functional problems on your user experience and ultimately the impact on your business.

dynaTrace automatically traces every single visitor and all their actions on your site. With this information dynaTrace understands the visitor’s environment (mobile/desktop browser, Dial-Up or High-Speed, geographical location, e.g. US or Asia), whether visitors make it to through your Landing Pages and where along the click path visitors lose interest and fail to convert.
Since dynaTrace knows where visitors bounce off your site it allows you to optimize these pages to improve conversation rates.
dynaTrace correlates page performance as well as functional problems to conversion and bounce rates. This allows you to focus on pages that are slow and pages that fail due to functional problems.

With its application-centric and transaction-centric view, dynaTrace promotes the application and business transactions to first class citizens. By examining the application and its performance first, before anything else, it can easily monitor environments with many 1000s of JVMs.
dynaTrace automatically captures each and every distributed transaction. It automatically captures the JVM/CLR/Host metrics of each participating node and correlates them to the passing transaction for later analysis.
These approaches enable you to easily monitor even the largest environments and still have all the data available in an easy intuitive form for analysis.

dynaTrace Collectors are the key architectural element to support globally distributed environments. Collectors service individual data centers and send this information securely to the dynaTrace Server.
dynaTrace supports a large list of commonly used service architectures such as Web Services, .NET WCF, RMI, JMS, Message Queues, …
dynaTrace supports the following ESBs: Oracle Service Bus, Fuse ESB, Oracle SOA Suite, SOA Software Service Manager and Apache OpenEJB.
dynaTrace's patented PurePath Technology can trace a distributed call through your globally distributed SOA environment end-to-end. Only by having a full end-to-end trace is it possible to monitor SOA transactions from a business point of view.

dynaTrace fully supports and is perfectly suited for cloud environments because of its transaction-centric monitoring approach. All agents are automatically discovered and mapped into the correct system. This enables dynaTrace to monitor large scale cloud environments from a business and application level and identify both application and cloud issues and their impact on your services.
dynaTrace fully supports EC2. For more information see the ( HowTo Guide)
dynaTrace fully supports VMWare and VMWare-based clouds.
dynaTrace can help you analyze your applications in Microsoft's Azure as well. (Learn How)

dynaTrace automatically discovers hotspots in every portion of your application. New and previously unknown technologies are automatically covered by this mechanism.
dynaTrace has an easy to use, browse and click based sensor system that enables it to be easily extended to add deep insight to any new Java and .NET based technology—without any special code support. For example, we can trace Big Data solutions end-to-end, add Business Transactions to visualize the different forms of operations, even enable deep dive diagnostics based on access patterns that rely on internal data like read/written rows, cache hit rate and amount of network calls. Also read the dynaTrace blog, Cassandra Write Performance – A Quick Look Inside

dynaTrace provides automation interfaces which are key for continuous integration. It integrates with the most common Build and Testing Environments including Hudson, Maven, Ant, MSBuild, Team Foundation Server, Selenium/WebDriver, JUnit, NUnit, HP Load Runner and SilkPerformer.
Agile practices demand continuous verification of functionality to identify regressions as soon as they occur. dynaTrace extends this process by verifying performance and architectural regressions.
With dynaTrace's Deep-Dive Browser Diagnostics capabilities one can also identify regressions in Web 2.0 Applications such as slow JavaScript Code, problematic AJAX calls or inefficient Browser Rendering.

dynaTrace provides a large set of Out-of-the-Box Dashboards that can either be viewed in the dynaTrace Client, in your browser, as PDF Report or even as an Excel file.
Custom dashboards can easily be created using Drag & Drop technology.
Every dashboard has a unique URL that can be shared with other stakeholders with a single click.