What is our primary use case?
Application process monitoring, and user-experience monitoring.
It's used for monitoring inside of applications, like JVM thread-level applications, plus the user. I monitor usage statistics, performance statistics, plus the user satisfaction level.
How has it helped my organization?
We got to know which modules of my application are used more by the customers, as well as which modules of my application are very slow. It has helped with my overall my performance statistics.
There are different phases of product development. In the testing phase I was not able to determine if things were going fine or what was going wrong in my application. But using this tool, I got to know - even before the customer got to tell me - this or that particular module was not working.
Using this tool, I got to know the moment he got a page-out error on his screen. It told me, for example, a person in the US is facing this particular issue. Because the alert came, we worked on it, we resolved it and things were easy.
The moment he called us we just said, "Yeah, we have already acknowledged that issue and we have resolved it. You can just try it again." The person was happy. The customer's satisfaction has improved, overall.
We also got know many internal code bugs which could have caused memory leaks or other issues which we were not able to catch during that development phase.
We also got to know, on the network level, where the latencies were. If you go via Google, what Google says is that you have a number of things on which you should measure your performance. One is if there's an error or not. Dynatrace tells you whether is an error or not. Second is saturation, whether something is getting saturated. You should be aware of what is getting saturated. Dynatrace even tells you that. The third is if there is a latency. Network latency is also told to me by Dynatrace. So these are three things I got out of Dynatrace.
What is most valuable?
The performance features are most important ones. That would include reporting, you get reports on your performance data.
The next is the alerting feature. For example, my applications are going out on some thresholds. So I get alerts, according to the thresholds I set. I get that data via emails as notifications. That is another feature that we primarily use and we like from Dynatrace.
Also, it's easy to use. The usability is better in Dynatrace.
The fourth thing is: my customer is facing some issue. The linking is very good in Dynatrace. What happens in other monitoring tools is the linking is not proper. In those solutions, a person has to manually link many of the layers and what is happening in them, while in Dynatrace you get that from the very first visit. For example, if a person is visiting your website, from there it will traverse you to the end. If the application is a Java application, it will traverse you there, to the Method level. So that linking and traversing is better in Dynatrace.
What needs improvement?
If Dynatrace is capturing everything in your application, it has to "sense" that information, and that sensing needs sensors which we have to include in our applications. The more you apply sensors - the more details you want - the more you have to increase the level of sensing. If I increase the level of sensing, my application's performance goes down, because something is there that is, again and again, checking each and every thing in the application. So that load on the applications increases.
So, many times my applications used to crash because Dynatrace was working on them. So that was a negative point. We had to remove some sensing; either we had to reduce the sensing or we had to remove Dynatrace immediately. So that is one thing we don't like about Dynatrace.
And the second is dashboarding. The dashboarding in Dynatrace is not very good. We have used other monitoring tools like AppDynamics. We are also using AppDynamics for some of our products. If I compare Dynatrace with those monitoring tools, the dashboarding is not as good. If I have to create a dashboard it takes me time, the experience is not that good. The automatically generated reports are good, but the dashboarding was something we were expecting but did not get.
Also, sometimes it happens that we are not able to capture things. For example, if a person is logged in from India, from the city of Mumbai, and is using a Chrome browser, and his email ID is firstname.lastname@example.org. But what happens is, Dynatrace just fetches two pieces of the information, not all of it. Sometimes it gets it all, sometimes it doesn't. So that also came into picture.
The last and the most important, which we did not like about Dynatrace - and that's why we switched to other monitoring tools recently - was the support. We were not able to get proper, good support from Dynatrace. We had to raise a lot of tickets and then, one fine day their people would eventually come around and resolve the issues. The input we had to give them was very high. Support was very bad for Dynatrace, especially in the India region.
For how long have I used the solution?
One to three years.
What do I think about the stability of the solution?
When you go for a monitoring tool, there are two types. One is SaaS. The SaaS version means that the monitoring tool is not deployed on your servers, they are deployed on their servers. The second one is on-premise.
We were using the on-premise and we were using very good servers for the Dynatrace deployment. Stability didn't come into picture. The version that they gave us was very stable. There were was no code bugs. Actually, there were some, but those were fixed immediately.
As far as on-premise was concerned it was fine.
What do I think about the scalability of the solution?
Scalability was fine. We were using around 100 application licenses, 100 Java licenses. On that scale it was able to handle everything. There was no reason to scale it up. I'm not sure if it could be scaled up to 500 or 600 licenses. We didn't do that. It was able to handle the load in the one tier.
How are customer service and technical support?
I rate tech support very low. The technical support was not that good. They were not very attentive whenever there were issues, even the critical ones. We were not able to find the proper support from their end so I don't rate it very well. A one or a two out of five.
Which solution did I use previously and why did I switch?
Initially, we were using New Relic. But, after that we switched to Dynatrace because of the amount of functionality, the amount of troubleshooting it was giving us was more. That's why we shifted from New Relic to Dynatrace.
But once we saw the negative points of Dynatrace, we recently shifted from Dynatrace to AppDynamics. We are in a process of shifting all applications from Dynatrace to AppDynamics.
How was the initial setup?
It was straightforward. It was was easy to add up. It was not complex at all, while the other monitoring tools are more complex than Dynatrace. Dynatrace was not that bad.
What's my experience with pricing, setup cost, and licensing?
The licensing for Dynatrace is high. If you want to go for monitoring solutions, then why Dynatrace? If you have a particular budget, you can go for many other monitoring tools - apart from Dynatrace - and they can help you more and give more data than Dynatrace can.
And secondly Dynatrace also comes with a lot of issues, which I have mentioned elsewhere in this review, which can easily be rectified using other tools. It's not worth the money that you spend for Dynatrace.
Which other solutions did I evaluate?
Initially we were using New Relic itself. Dynatrace was one thing that we were evaluating. ManagingEngine was a new one at that time from what I recall but, I'm not sure which ones were part of the evaluation, because I was not totally a part of the PoC.
Apart from that, if you want to use on a system level, you can use Nagios, that's freeware. It's also good but, again, it is just a system monitoring tool. It's not an APM. So if you wanted to go for APM, then only New Relic. It was the one competitor for Dynatrace, at that time.
What other advice do I have?
In terms of implementation, it's quite easy, now that there are many automation tools. So just integrate Dynatrace with the automation configuration tools. Just ask Dynatrace which integrations it has, for example Chef, or Puppet. If you integrate, the configuration will be easy.
Also, the configuration needs to be standard. The standards should be set initially. There should be a standard protocol; that needs to there. If that's not there, then issues may arise later on. These are some things which are advisable when you work with Dynatrace.
I would rate Dynatrace a six out of 10. When I consider all the negatives plus the positive points which I have already discussed, I end up at six, including the licensing and everything.