What is our primary use case?
Our primary use case for both Dynatrace and AppDynamics is for application performance monitoring (APM). The main reason for having application performance monitoring is, when we see something is running slowly, we can immediately look to see where the issue is at before our systems crash on us. So, one of the major roles it plays for us is the ability to keep our system performing in peak shape.
Our ability to see issues coming, then quickly isolate and correct problems was our main use of Dynatrace. We are not there yet with AppDynamics. It has been ten months, and we are still spinning our wheels trying to set it up and figure out how it works.
How has it helped my organization?
When collecting data with Dynatrace, we saw every single transaction that happened in real-time. Whereas, with AppDynamics, they take snapshots, and we only see a tenth of the information that we did with Dynatrace. While the information is there, if an issue with an application happened in-between snapshots, it would not be readily identifiable. You would have to go hunt and peck for it. We don't have time for that.
What is most valuable?
Using Dynatrace, we collected application metrics within three hours, in most cases. The majority of our triage were within three hours, then we were able to discover the root cause of issues.
With Dynatrace today, you have a single agent. You stick it on a server, and it doesn't matter if it's Linux, Windows, etc. It is a single agent and executable. You run it, and it injects itself into your collecting data. This is compared with AppDynamics, which on some of servers, we have had to install as many as four different agents and configure them all individually, trying to collect the same type of information.
The dashboarding for Dynatrace is ten times easier to set up and has more options of what you can put on it, especially if you are in a single payment class environment.
What needs improvement?
With Dynatrace in our environment, the managed server required root access to run. As a government agency with tight security, this has been an audit concern for us. A major area of improvement for Dynatrace would be to make it so the program does not need root access to perform. AppDynamics does not require root access to the servers. Once they are set up and configured, they can set their end run without root access.
The number one area of improvement for AppDynamics is to simplify their agent install. Instead of having four or five different agents to get all the different things that you need with different pieces of information, they need to figure out how to put theirs into a single agent, like Dynatrace has done.
We have not found AppDynamics in our environment useful at all. We are struggling to try and make it work. AppDynamics is for applications that are static. In our government agency, we are too dynamic. Everything is changing constantly, and AppDynamics does not work in this type of environment.
For how long have I used the solution?
More than five years.
What do I think about the stability of the solution?
The stability for Dynatrace is probably 85 percent. We had enough issues where one of the services would just stop running, then we would have to restart it. Not very often, but it happened.
With AppDynamics we haven't been able to use the system long enough to determine its stability.
What do I think about the scalability of the solution?
Dynatrace is very scalable. We grew it over the 12 years that we had it. So, that has proven that it is scalable.
AppDynamics is scalable as the environment grows. However, we are still in the product's infancy, so we haven't seen this happen yet.
A company with a single application over multiple locations, like a retailer, but only needs to worry about one application and monitoring it, this is the perfect fit for AppDynamics. If you have an organization with more than 40, you definitely want Dynatrace. AppDynamics is not a viable product for an organization with lots of applications.
How are customer service and technical support?
On a scale of one to ten, my experience working with the AppDynamics onsite people and offsite support is maybe a five. I feel that they don't want to take responsibility for the areas that AppDynamics is lacking in. When things don't work in which they were sold, then they want to tell us that it is our environment more than their application not functioning correctly.
Whenever we had issues with Dynatrace, you could get Dynatrace support on the phone, and they were all over it. They would get into our machines, then take screenshots or look at the performance of the systems while it was running. They wrote custom patches to help us resolve issues that we had. I would rate the support that we receive from Dynatrace as a ten out of ten.
Which solution did I use previously and why did I switch?
We have been using Dynatrace since 2007 and AppDynamics for just about a year. We have not been using them concurrently. Dynatrace was not renewed, and management decided that we would use AppDynamics. This decision was beyond my control.
How was the initial setup?
With Dynatrace, the installation and setup were a piece of cake. It could be accomplished usually within fifteen minutes, and definitely, within a half hour of deciding to do it.
A big difference that we found between the two vendor is in setting the system up and getting them ready for production. With the latest version of Dynatrace, it took three days and we had it in production. We are still trying to get AppDynamics in production since last May.
What about the implementation team?
With our Dynatrace system, it required three servers and the program when we installed it on the servers. It was straightforward. You just clicked, clicked, and clicked as it went through the setup, then you were done.
With AppDynamics, we are now on eight servers to make it function like it should. We are ten months into it now, and we're still trying to get it right. We have AppDynamics folks onsite to help us with it. It is just difficult to implement. There is so much to it.
I sat down with one of their architects to decide how many servers that we needed, how they need to be spec'd out, etc., and that is what we built. Then, when we stood it up, and it didn't work, some other personnel from the AppDynamics team came in to look at it. They said, "None of this is correct. You are on the wrong this and that."
Thus, we did not have consistent information and support from AppDynamics when implementing the system. However, it is up and running now.
What's my experience with pricing, setup cost, and licensing?
Our annual costs were about the same for both AppDynamics and Dynatrace.
What other advice do I have?
When learning Dynatrace, we brought in Dynatrace people to come onsite and take my team through a week long training. We did that two or three different times. They offered this type of training. They also have online training out on their community that I could set up for my team members. The effectiveness of that training was about 75 percent.
With AppDynamics, they have provided some online training. The take away from it (from my team) has been maybe 10 to 15 percent. The training is geared more towards sales than using the product for what it was intended. It boasts the features and selling points of the AppDynamics product instead of the ins and outs of how to use it once it has been installed in our environment.
I would definitely recommend Dynatrace. I have the benefit of having used it for so many years. It takes less infrastructure to set it up initially. It's a single agent engine. You just set the agent up and run it, then it configures itself. It goes out and finds all your processes with everything that's running, configuring itself. The simplicity of the infrastructure and simplicity of setting it up, then actually using it, along with setting up your dashboards to monitor your metrics is much better. There are more features than the AppDynamics dashboarding.
I would rate it Dynatrace as a ten out of ten.
At the point of where we're at with our AppDynamics experience, I would rate it as a five out of ten.