Dynatrace Reviews

This solution is used for:

• Integrating the diagnostics for all of my team members with the same point of view through Dynatrace link.
• System integration for automatic remediation; to launch Ansible Tower jobs in response to problems or events.
• User experience tracking based on session behavior.

We have a DevOps team distributed between several offices and we need to keep track of every microservice evolution. Dynatrace helps us to do that, keeping the same dashboards and same links for problems in our production environment, or profiling at staging. 

I think one of the most valuable features is the Dynatrace API, for both metrics and configuration. With that API, we were able to launch automatic tasks or jobs in response to a specific condition. We also have the option to open a JIRA ticket to keep track of a particular issue. 

I think Dynatrace needs improvements with respect to reporting; not just performance, but the business-level reports. The navigation can be improved because when you press the back key, sometimes you lose the time frame. Also, when you are in a problem description and want to leave, it is hard to do.

While working for an MSP, these customers' use cases vary - in most cases, we must do the fault domain to identify who is responsible in order to fix the root cause.

One use case: A customer blamed our data center, stating it caused broken sessions for their Citrix users. After looking at the network traffic data with NAM (ex. DCRUM), it was easy to identify issues within ISPs at Asia, and not in our data center here in Europe.    

All of them which measures performance and availability. If you don't measure, you don't know. Looking at the data patterns and trends, you can also find out abnormalities in the whole ecosystem, not just in the target application.

Reporting and dashboards could be better, compared with competitors. However, Dynatrace has excellent API for data mining but requires extra reporting tool.  

Product is very high tech and advanced on the one hand, and very easy to implement and maintain on the other. 

We are not blind anymore with our digital services performance; no more fire fighting. IT Ops is now proactive and collaborating with our development teams.  

• Code level visibility
• Smartscape: Modeling of the whole IT environment. 
• User experience analytics

• Product reporting still needs improvement.
• Session export for offline analysis, like it was in AppMon, would be also nice. 

Time to value was surprisingly fast.

Releasing a new product and going from an AS/400 to a Microsoft environment. 

We obtained a better insight into our environment and consolidated a lot of our old apps into one app. 

We are still implementing Dynatrace. We are in the process of a PoC to discover why our sync test failed. It also gave us the cause of what was at fault.

PurePath gave our developers some tools that they did not know existed, and they gained a faster, more robust use case.

It needs education and training to ensure you get the full value of your purchase. Maybe add in a certification for Dynatrace.

It's performing well. We typically use it for its intended purpose: application monitoring, identifying trends of release-over-release of our solutions, of average response time, impacts, server metrics. 

Specifically, for the subteams that I'm on, it's very much allowed us to get real insights into the impacts of our changes. It's allowed us to see pre- and post-metrics and be able to provide dashboards for our business. It's been extremely valuable.

One aspect of development is the concept of continuous improvements. There are key screens in our applications. We get identification, through Dynatrace on its own, that these are our top ten slow preforming screens. They choose to invest in optimizing those, and then we can show them the outcome through all the dashboards.

In the AppMon, offering, currently, would be just the PurePath analysis, being able to deep-dive into call chains.

The session replay. That's probably the biggest. I think that's the struggle right now, the ability to reproduce the customer's behavior. "Oh, I had a spinning wheel," or "I observed a error." And you're wondering, "Okay. How? How did you do that?" and they say, "I don't remember." Being able to replay exactly, the exact screen movements and everything, it's very indicative of a good feature.

One of my key focus areas does deal with performance testing, and Andy, here at the Perform 2018 conference, had a good session on performance testing. But it was a lot of utilizing a custom thing he built and "hooks" there, but they should make hooks into some of the more modern performance testing tools a little easier. I think that would go a long way.

We actually experimented with Managed initially, very early in its introduction, and it very much came off as a beta product, because some of the core capabilities were just generally throwing errors. But in what I've seen in the demonstrations here at the Perform 2018 conference, it has gotten a lot more polished. The AppMon part itsself is also better polished. So, I think it's the nature of trying to rewrite from the ground up.

I know, at least from an AppMon perspective, there's very much a limitation on memory and things of that nature. I think with the move towards the Managed product that they have some better opportunities there from a scaling perspective.

I, myself, have not directly used technical support. We have a team that actually focuses on directly tooling. They're typically a go-between with support.

Custom, home-brew.

When it comes to the nature of digital complexity, I think AI is significant when it comes to IT's ability to scale in the cloud and manage performance problems, especially in regards to shifting the need for manual observations, and in terms of identifying system degradation, etc. I think it's extremely valuable in terms of being able to anticipate potential issues, as opposed to the typical reactive identification of issues. I'd rather an AI system find it before a customer communicates such.

Regarding one solution that could provide real answers, as opposed to just top-level data, I don't think that's a possibility. Unfortunately, I don't think that there's always a one-solution-fits-all to any problem.

If a friend said he was looking to adopt an APM solution I would tell him, "Use what we have available in the enterprise, which is Dynatrace."

Java-based microservices (couple of dozen) monitoring and their dependencies (Oracle, MySQL, RabbitMQ, and Elasticsearch). Apart from monitoring (with alerts), I use the Dynatrace solution to investigate reported issues by clients.

It eases the investigation process (faster feedback loop with clients), better pro-activity on possible outages (in order to avoid outages), and eases the process of fine tuning allocation of resources to processes.

• Dashboards: Most critical information on a single view and in a very-customizable way.
• Alerts: Stays on top of probable issues.
• Services monitoring: Helps fine tune resource allocation (CPU and memory).

I would like to have more visibility into Python processes. Meanwhile, we are thinking about "fronting" our Python processes with a fully supported web server, such as Nginx.

We use it for proactive monitoring, user experience monitoring, analyzing load versus response time.

Very good for improving application performance. Gives us a view of the application, infrastructure, and users' experience with the application.

It identifies all the functional errors and slow APIs.

End-to-end monitoring could use more default dashboards.

Synthetic monitoring. Also, to be able to pinpoint issues on the web and mobile interfaces before they trigger a negative experience for the customer.

We're using Synthetic along with AppMon, and this has been very helpful in pinpointing errors and new trends to help with investigation.

Interactive charts and waterfalls in Synthetic, because they tell the story.

I also love the support we get through the forum and ticketing system.

It would be great to have Synthetic automatically retrieve what the customer sees on his side.