I need a comparison between AppInternals and Dynatrace APM offerings.
What are the pros and the cons of each solution, and why would you buy one rather than the other?
i don't 100% agree what @Scott mention.
1st - Please refer to the latest Magic Quadrant NPM & APM
Dynatrace is good in APM space but Riverbed is offering end to end monitoring in NPM & APM space and the latest acquisition Aternity expands Riverbed’s SteelCentral offerings up to the end users performance monitoring level wish i personally love it.
Traditional solutions base in DC to analyse packet and from the packet turn into end users performance monitoring however with the Riverbed Aternity is provide a real end point monitoring rather than monitoring in DC.
Back to the APM solutions, since the question is more on AppInternal & Dynatrace
the main differentiation is the language code , depending which applications that you require to monitor and what language code is the application using.
Dynatrace -Supports .NET, J2EE, PHP, Mainframe, Hadoop
Riverbed - Supports .NET and Java - thats all
Instruments .NET and Java Runtime environments to give you access to actual code
Others technical differentiation will post to you later
Riverbed is more of an NPM probe-based offering and only has a moderate level of application awareness. It is more aligned with Dynatrace’s DCRUM offering competitively. https://www.dynatrace.com/topics/performance-test/data-center-monitoring/
At some point our DCRUM will be integrated into the Dynatrace product, so one great thing coming is you will truly have a single platform/pane of glass for everyone down to the full network detail. For now, it is still a separate entity.
Although Riverbed offers strong NETWORK analytics, it is not going to provide the application layer detail and provide a holistic all in one offering like Dynatrace does with RUM, webchecks, cloud, containers, network, infrastructure. It also samples and doesn’t provide a high fidelity of data. Dynatrace looks at every transaction, so we offer gap-free visibility into performance, bottlenecks, and issues. This is also important if you want to understand user and performance trends and understand where to align resources and focus on areas of development. With Riverbed, you’ll be making a lot of assumptions based on their sampling/averages output.
It’s going to take you a month plus to evaluate it…minimum. It takes weeks to set it up and configure. And the cost of services to get installed and trained is costly. So ease of use with what you evaluated is not comparable. Dynatrace installs in 3 mins or less and our ROI is typically within 2-3 months.
Maintenance of Riverbed is huge. Dynatrace is automatic in providing releases and upgrades, zero maintenance. Riverbed is going to require taking things off line, doing the updates, and putting things back on line. Lots of manual effort.
Riverbed strongly emphasizes packet capture capabilities so it’s appealing to the network team. It captures and records everything from the wire and applies very light weight analytics. It’s a very reactive approach. If you want to apply the application layer information into the networking piece, you will be purchasing and managing multiple components instead of just one with Dynatrace.
Dynatrace is application centric however it does provide some network analytics in relation to the performance of the applications it’s monitoring; This is like comparing an apple to a banana.
So, for me, the top areas in which Dynatrace is better:
• Ease of use, zero maintenance
• Higher fidelity of data with Dynatrace-no sampling, aggregates or averages like you’d get from Riverbed.
• Quicker ROI and user adoption
• We show you every user, every app, everywhere. We provide gap free data from end-user, code, infrastructure, network. No blind spots, no samples, no averages. So no matter what device you are operating on, we’ll provide gap free visibility into performance.
• Zero manual configuration. Just install one agent per host, we monitor everything.
• AI gives you auto-everything. Automates, discovery, modelling, analysis, troubleshooting and stops you from having to figure this all out manually.
• Automated root cause analysis. Avoid alert storms, get one single notification.
• Seamless integration for cloud and containers.
Disclaimer: I work for Dynatrace. However, we work with many customers who have used both Dynatrace and AppInternals.
It should be noted up front that the below comparison is about the Dynatrace platform (formerly known as Ruxit), not our prior architecture solution AppMon. The Dynatrace platform is our primary product today.
To start, there are some fundamental differences between the two solutions. AppInternals is a traditional APM tool where you instrument your Java and .NET apps. It sees the user experience and looks at stack traces to see the application transaction flow.
The Dynatrace agent installs at the host level and automatically discovers and instruments everything running on the host. It is a single agent, regardless of the technologies being monitored. Because of this approach, Dynatrace is able to do full stack monitoring, including the end user experience, the application, and the underlying infrastructure, including log files.
Because of Dynatrace's host-based agent approach, implementation is extremely easy. Simply install an agent on each host, and the entire application environment is auto-discovered in a matter of minutes. There is no additional configuration required. It can be used as a SaaS offering or on-premise. The AppInternals agent must be configured to instrument each Java process, regardless of how many are on the host.
AppInternals supports Java and .NET only. Dynatrace additionally supports node.js, Python, PHP, Golang, and more.
Anomaly detection/root cause
This is one of the biggest areas of difference. Dynatrace employs an artificial intelligence engine to automatically set baselines and performance thresholds, going beyond simple averages and standard deviations. Dynatrace then automatically detects anomalies, understands the business impact, analyzes the dependencies involved, and identifies the root cause. The AI engine is deterministic, so you get true causation as opposed to correlation that AppInternals and other tools provide. AppInternals works more like a traditional monitoring tool. It shows you transactions that are performing slowly and lets you drill into the problem in various ways, but ultimately it leaves the root cause identification to the user.
Digital experience monitoring
For monitoring the user experience, Dynatrace goes much deeper than AppInternals, looking at additional metrics beyond simple page load times to give you "visually complete" times, more in line with what the user actually sees. Riverbed’s ARX product is decent for a network view of legacy applications, but wasn’t built for modern apps. Dynatrace also has synthetic transactions built in, while Riverbed’s synthetic monitoring is in a separate product.
AppInternals is primarily used for traditional app environments due to the instrumentation approach and limited language support. The Dynatrace platform was actually designed for cloud native environments. It automatically detects and instruments containers, provides full stack visibility in Cloud Foundry and OpenShift environments, and is available for purchase directly from the AWS and Azure marketplaces.
Lastly, I'll just say that Dynatrace is really easy to try if you want to check it out. https://www.dynatrace.com/trial/
No doubt Dynatrace is very good and efficiently work. AppInternals is also good one but mail it works for Network guys. If you are interested in Network analysis related to application ( need manual effort). If you are looking for a good solution you must go for Dynatrace or AppDynamics.
We are using both Dynatrace and AppDynamics depending on the application and Business demand.
Similarly Optier, another great technology innovator in APM, pioneered the use of advanced analytics in the space, didn’t suffer from vision or technology. The issues were once again the leadership outside of the technical parts of the organization. Having raised well over $100m in the last 9 years the company never kept pace with the changes in the market. Optier has finally closed their doors as of yesterday, really sad to see that the transformation from on premise heavy enterprise software to SaaS was not happening fast enough to fix the cash situation. Hopefully someone will acquire some of these great assets and possibly see the transformation through.
I am not familiar with app internals. We have adopted Dynatrace. Its sadly not as simple as Scott makes out though. In the case of app like BizTalk, you need to map the app pools and IIS services. One agent doesn't do everything, there are agents for each of the apps and a DB agent will not give you all the host metrics.
That said, it is a great tool and does all we need it to do for some of the reasons mentioned above around one tool that covers EUM, APM and a Network Probe and the ability to easily dril down to the most granular detail.
I am not familiar with these two anymore. I'm sure you can Google around as well as I can for general impressions, pricing, etc.
I'll offer a bit of process guidance, though: Read the Release Notes for the latest version of each product. That will indicate what each company finds significant, and you can judge how well that aligns with the needs of your platform and your mgmt.
I'm comparing AppDynamics and Dynatrace for a company with over 5K employees. What are the pros and cons of each pproduct?