What is our primary use case?
My primary use case is troubleshooting production issues to see which endpoints are mostly calling and get some logs about them.
It's performing well. We use AppMon, and whenever any production issues occurs, we get a time window from the customer explaining, "This time it is a software failure."
We actually did some training on the AppMon UI, then it almost never failed on us. We could trace back exactly where the issue occurred. We came to the Dynatrace event to understand more about it, and see how we could learn more to perform our troubleshooting better.
How has it helped my organization?
It used to take a lot of time to troubleshoot. Now, with Dynatrace, we can actually see the logs anytime we want. I can just find the problem. It has improved performance from a time perspective.
What is most valuable?
The most valuable feature the solution offers right now is the PurePath. When we see a web request, and something failing, we can drill down using PurePath. Then, on PurePath, we can get to the database SQL statement. So, it is a very cool thing. We have our logs and all the stuff, but this is an area that stands out.
What needs improvement?
The AI thing is a boost, but it is not in AppMon. It is in Dynatrace. If AppMon could also incorporate it, that would be great.
The main thing is more about the dashboard. Currently, it does not keep the snapshot data. It only keeps it for a very short duration. Because of that reason, we cannot get more reports. If we can somehow option a way to preserve the data and keep it for a longer time, or have that feature, that would be good.
What do I think about the stability of the solution?
It is always requiring us to update the Dynatrace client. There are some issues with Dynatrace. Many times I thought there was a production night deployment and we would do a smoke test. Therefore, I started Dynatrace, and suddenly it says your client needs to be updated. Then I had to go do that two or three times. So, there are some glitches in it.
In the future, Dynatrace should actually try to resolve those issues with AppMon. The new Dynatrace version, I think will not have these issues, because it will be only browser based. So, there will be no need to install anything on our machines.
What do I think about the scalability of the solution?
I sometimes see issues with the loading of the data. If we give you criteria, like you want to see the last 15 days of data, it has a lot of requests in it. To refresh, it will take a lot of time. Therefore, we actually narrow it down to a particular instance or event.
My main purpose is to troubleshooting issues, so this way I know exactly what time it happened, then I can just narrow it down to that. However, if you want to see a month's data, it keeps on spinning. Here is an improvement which needs to happen, which is the case with all applications or tools. There is a lot of data, and either we have to change the way we are logging or the application needs to be enhanced.
How are customer service and technical support?
For some reason, we wanted to create a report about what is the maximum used end point or the maximum called end point. Also, we wanted to create a certain dashboard at that time. Myself, with a DevOps engineer, we called the technical support team, and they helped us create that graph. We got it, actually.
Which solution did I use previously and why did I switch?
We have used production logging. We have our own custom logging system, where we read the logs and transfer to the database. We have a nice UI for it too, but it does not scale well. It can read the log, but it can read only those logs which we added.
Most of the time it works, but it is very slow. If we want to trace back, for example, to find out what happened in the previous week, we cannot get that data.
How was the initial setup?
The DevOps engineer worked extensively on it. I do the troubleshooting.
What was our ROI?
What other advice do I have?
It is the best solution that I have seen so far.
AI is like a new feature or benefits, and it is a cool thing. We have not tried it, but I really like to see it working. It is a great program, and it automatically makes a trendline baseline. Whenever something goes wrong, it can send an alert. There is also a web check feature.
Currently, we do not have a baseline. If AI is there, it can see the trend. Based on the trend, it can send notifications. It is also integrated with various platforms, like social platforms, so we can also get alerts from there.
Most important criteria when selecting an APM solution:
We look mainly for how it will scale and what are the features currently available.
I provide this information to the decision-makers.
Disclosure: IT Central Station contacted the reviewer to collect the review and to validate authenticity. The reviewer was referred by the vendor, but the review is not subject to editing or approval by the vendor.
Feb 21 2018