Dynatrace Platform Subscription (DPS)

My main use case for Dynatrace is that we find it to be a good tool for working with establishing analytic models for our customers' front ends or providing dashboarding for customers in relationship to their operations.

Regarding my main use case with Dynatrace, it helps with the ability to formulate a single pane of glass in regards to working with automation and operations, including DevOps into an API or an MQTT model, which allows you to monitor the orchestration of how updates are being managed and handled, how effectively those deployments are progressing time-wise and personnel-wise, and to expedite the course of upgrades and so on without introducing infrastructure overhead while allowing everyone to see what is happening at once.

Dynatrace has positively impacted my organization because reporting is ongoing and constantly evolving to adapt to the new technologies that are being introduced, such as AI and other standard protocols of reporting and metadata. The key advantage here is that Dynatrace has a good position in this space by being flexible enough to accept variations and bring in unknown footprints of AI and APIs that will introduce new data sources, giving it a chance to be very flexible in receiving those new standards with very little overhead. Not everything can come in quickly, but many things have, and Dynatrace is particularly good at that.

Specific outcomes or metrics that show how Dynatrace has helped my organization include its helpfulness with RCA reporting for incident response. The key details here are that you have to understand when something does happen, whether it is a fault on the service side or a fault on the customer's side, and understand the brief hiccups that cause outages and impacts. The RCA reporting process for determining the cause, conducting a causal analysis, recovery causal analysis, and measuring the effectiveness of that response to justify and explain when things happen, specifically in the incident report response, has been valuable. When this is done holistically within the reporting analytics tool, it can greatly reduce the investigation time period to completion. The key detail here is that when an RCA investigator report is initiated, all of the top skill sets are engaged in that process, which represents a great deal of money and time that is lost because all team members, in their own capacity and relationship to the investigation, have to stop, do the analysis, reach a conclusion, do a follow-up corrective action item planning model, and schedule an ETA to get that accomplished. Those are very important details that take a lot of time, and Dynatrace helps reduce that time frame.

The best features Dynatrace offers are somewhat subjective as it does not do anything functionally unique compared to others. However, it does work very well with integrating into other environments or working in tandem with other environments. A reporting tool like Atlassian allows you to generate analytics reports for leadership structures and documentation in regards to that kind of support. This is valued because when we work with tools such as DataDog or New Relic and other major monitoring platforms, we can send all of that information into a very condensed leadership equivalency level reporting matrix that the leadership of your company can actually read, understand, and take steps with and act on because of the way that the tool comes in reporting-wise effectively up into the consolidated enterprise reporting metric systems, such as Atlassian.

Regarding the features, Dynatrace works in two different environments. Its siloed version, the older version, is usually hosted on local infrastructure, whereas its new possible format, the SaaS version of Dynatrace, offers additional value as well as beginning to solidify the AI models that will be under the hood for the Dynatrace engine SaaS model, which adds value opposed to looking at something the server version of the older release of Dynatrace. The value is that SaaS gives you that level of freedom, and of course, the MRU and SLA agreements that you have in place to support Dynatrace SaaS services give you an added level of security that your own siloed environments do not offer.

Dynatrace can be improved by focusing on the model because when you say single pane of glass, it means different things for different people. When you have a single pane of glass for your customer base, and let's say you have 190 customers, technically speaking, that is 190 dashboards that exist and at least a template format with some optimized additional monitoring requirements added to them. However, that is not the same as a single pane of glass for infrastructure, which Dynatrace also serves. There is also a single pane of glass for executive leadership of IT operations and environments, including automation and the SRE Director of Operations role. This is not the same as a single pane of glass for executive management, such as the CIO and CEO, which require a completely different type of set of tools. The preparation of FinOps for finance, SecOps for security, and for automation and operational roles needs to address the company requirements specifically in regards to grid systems or distribution systems. Each of these will have its own version of a single pane of glass, and that is something that Dynatrace needs to be more flexible on going forward.

I would also like to note that Dynatrace needs to have the ability to import additional types of imports beyond standard containers, images, and animations. Animation value does carry significance here and it would be valuable to be able to add that because animation is an initiator. Most companies that use animation in their environments tend to get faster responses and response times as opposed to simply using red, green, and yellow. This can add value to that discussion.

Other improvements needed for Dynatrace include understanding that one tool usually cannot meet all the requirements you need. Being aware of that means being more pragmatic about your approach. By introducing an additional tool or two to cover all the aspects and attributes, technology is always evolving, and you may find a lesser-level vendor who can monitor something newer than what your overall environment is monitoring with something such as Dynatrace. Dynatrace is not really a bad tool when working with other similar services, such as DataDog, New Relic, or Checkmk. Dispersing your resources in multiple baskets provides more reliability, sustainability, and better control of cost factors because you have an easier way of gauging costs between vendors versus services. The key thing is to ensure you have internal and external viewpoints and multiple potential tools such as Dynatrace to help you achieve 100 percent accountability and reliability.

I have been using monitoring tools for about six years, which includes Dynatrace.

Dynatrace is stable.

Dynatrace's scalability is to be determined since we are on SaaS and we have not had a scale event yet, but that is something we will be looking into as we develop the business continuity model.

Customer support for Dynatrace is good, and I have no issues. They have provided good services for our needs. I would rate the customer support an eight.

I previously used many solutions out there. The biggest driving force for me has been getting the most value for the investment for compatibility, integrability, and serviceability to a single pane of glass based on my requirements. This is an evolutionary thing, and while we may be using Dynatrace now, we might switch to something else such as New Relic or another type of tool or even progress to an open-source tool to better control costs in the future. However, all in all, Dynatrace is doing a good job, and we are using it for that reason.

I have seen a return on investment, as my example of ROI is that using a tool such as Dynatrace is an investment that reinforces the customer's desire to want to stay with us. This is crucial because as you integrate these processes, you also want to integrate the investigation process, what we call the RCA process. It is important to move forward with the least amount of time and man hours as necessary while keeping your involvement with other employees as minimal as possible. This enables you to complete the investigation as soon as possible, ultimately saving in operational overhead, productivity, and personnel resources.

Regarding my experience with pricing, setup cost, and licensing, the setup costs and licensing were not a problem for Dynatrace and have not been a problem for most monitoring services. The key issues that were upfront were information processing, data source processing, and housing. We understand that Dynatrace is in business to be in business, but these areas specifically are where you have runaway costs. Your initial investment in licensing costs pales in comparison to the question of why you would charge a license fee and usage for a tool if you are going to be three times higher in processing the data you are sending via the agents. It is really important to know that if you come across costs on the actual operational side, you can use other vendors to help offset that to a small degree, ensuring they do not have a cost covering from something such as data processing or data storage. Alternatively, you could use your infrastructure and local data centers to have an Amazon repository for logs and alerts to get those costs under control. Cost control is very important when using these tools.

I evaluated other options before choosing Dynatrace.

Regarding Dynatrace's AI capabilities, I think its governance and security are only as good as the leadership and the infosec overhead involved. Dynatrace has good templates to meet the AI compatible capabilities, but it always comes down to the human element in the environment. Proper correspondence and response are required at the human level to fully bring the power and capacity of the AI capabilities into its full effect.

The accuracy and reliability of output from Dynatrace are based on two very critical details. One is the contiguous articulation of input that comes in from all sources without interruption. Secondly, an overall average gives you the average reliability with said accuracy. This basically means that if you are tracking everything effectively, when you do have hiccups or impacts, using those capabilities will help identify where you have what we call black holes. The black holes deregulate and devalue the nature of the reliability of this kind of solution because the processing agent configurations may not be monitoring everything necessary, and the importance of the AI capabilities is to ensure they are identified, quantified, stood up, and put in place to help boost accuracy and reliability. That does not always happen and it is human-dependent and must be monitored to make sure that it is done correctly.

Dynatrace is deployed in my organization in both on-premises and some in public cloud that is also deployed with two other monitoring tools, both of which are SaaS-based. We use a SaaS provider for the public cloud deployment.

My advice to others looking into using Dynatrace is to document every application and every infrastructure resource that you have and ensure you place those into a central repository or a wiki. This will allow you to have that information in place to do a comparative design model and see if you even need Dynatrace in the first place. Small to mid-sized companies may or may not have a need for Dynatrace, but this is the most effective way to approach it. I give this review an overall rating of nine.

My main use case for Dynatrace is observability, monitoring, insights, performance, variation, dashboarding, services, traces, validation, cross-team collaboration, reporting, and many more.

As a senior performance analyst, I look into Dynatrace in terms of the observability parts, and mostly, there are multiple dashboards that have been created for multiple product groups from front-end layer to a back-end layer. I keep monitoring their performances, their failure, and their anomalies that have been detected through the problems page, alerting if it has been sent to the right team, and investigation on why these issues do occur, how are they recovered, and who would be the right team to contact. One of the newest cases would be when we had extremely high load on our application, and we were unable to figure out what was causing it. We looked into Dynatrace and noticed Apdex being down, high load days, and request errors being increased, and investigated to identify that it was the bots that got increased, but they were all coming as real users. We tried to investigate their user agents, we tried to investigate their IPs, connected through a bot manager system and CDN, and then mitigated the issue through content protector. We were able to improve our performance, our Apdex, and we did reduce our overall load on the site.

In terms of my main use case and other ways my team uses Dynatrace, I would say teams have collaborated. DevOps uses it for administrative purposes, including me, and then we have developers use it for investigative purposes on our lower environments like QA and Dev. We have a separate tenant specifically only for China users because they are restricted through firewalls and all of those things, so the servers are all set up within China, Tencent, providing a specific tenant only for them. These are most cases.

Dynatrace has impacted my organization positively in many ways because we have been using it for almost eight to nine years now. We have used Dynatrace day in and day out for performance improvements. The recent example I would share is from our front-end team, which used it in daily investigations of how our web vitals, including Google Web Vitals, are performing in terms of LCP, INP, DOM interactive, Time to First Byte, and other web vital metrics. We were able to figure out that some of our images or some of our JS scripts were the primary cause of these being down, and we were able to use Cloud and other agents through Dynatrace to find the slowest performing components, fix them, and we have been able to see those improvements. Our site is way better compared to three months ago.

To measure those improvements, we see that our LCP actually improved by about 42%, trending around 5.3 seconds on a median range, which we improved to around three seconds range. In the same phase, we also improved our TTFB by somewhere around 40-42% and DOM interactive by somewhere around 40-45%. We are also trying to improve on multiple other categories, not only on overall site performance but probably homepage performance, content performance, product page performance, search page performance, and others. These improvements are in progress.

The best features Dynatrace offers are many. I would say from dashboarding, alerting, integration with multiple applications, the new DQL features that come up with Dynatrace, the notebook feature, the workflows, the ready-made dashboard that Dynatrace provides, the page where you are able to track your license consumption, and the observability part where it provides the root cause for the incidents that appear.

Out of all these features, dashboarding has been a major contribution in my day-to-day work, looking at performance trends, failures, errors, and all of those things, and cross-verifying that with the problem detection that Dynatrace provides. Over the recent past, DQL has been a major help, especially for dashboarding purposes, investigative purposes, particularly through notebooks, since you do not have to run a dashboard always; you can just use very less license consumption and run a query to find out what the actual trend looks like or what the failure looks like and where exactly this is. Most of these are collaborated with Dynatrace Assist, AI Assist, and Cloud Assist through MCP servers, which has also been pretty helpful.

Certain configurations still require some navigation through multiple screens. Dynatrace offers many apps, but when it comes to training and documents or help from a point of view, it is very elaborative. You need to go find, search, look into the specific ones, and see how they are changing. Dynatrace currently offers both classic and Gen3 versions, so most of these documents are kind of updated with Gen3 while some users may be using the classic version, making these documents sometimes confusing regarding their availability for classic or Gen3 versions.

I would also like to add that the user interface has been helpful, allowing all the apps it connects to be accessible through a single stop or a button, where you can get all the apps and then look into them. So in terms of user interface, it is pretty helpful.

Service capabilities are great, user communities are great, the relationship with Dynatrace support, CSMs, or account managers has been great, consulting partnership is amazing, and the breadth of services is awesome.

Regarding Dynatrace's AI capabilities, I think its governance and security are handled well because until this date, there have been no complaints or issues we have come across. I agree that the governance has been taken care of, and all these CCNA, CCPA, and all PII level priorities are considered, so they have been good in that respect.

Regarding the accuracy and reliability of output of Dynatrace's AI capabilities, DTCTL has been great, and I would probably give it a percentage of somewhere around 80-85% on its accuracy. Through Dynatrace Assist, although it uses agent AI, there are times I have seen not satisfactory responses. So I would probably give it somewhere around 75%.

For others looking into using Dynatrace, all those examples I shared earlier in terms of pros have been great. I would say anyone really willing to purchase Dynatrace can use their test features for testing or POCs that Dynatrace would provide for approximately a month period. This would be helpful in understanding what the tool can provide and how beneficial it is for your organization. I rate Dynatrace a nine on a scale of one to ten.