Datadog vs. Dynatrace: a side-by-side comparison for 2022

Dynatrace and Datadog are among the most popular observability platforms available. Both offer many features integrated into one observability stack, but there are key differences.

We've deployed and tested both, and below, you'll find a side-by-side comparison of Datadog and Dynatrace for 2022.

I've decided to compare these tools based on the following criteria:

1. Platform functionality overview
2. Ease of integration
3. Onboarding, UI & UX
4. Incident management
5. Pricing
6. Third-party and community reviews

⚖️ Testing environment

I've deployed both tools on Ubuntu 22.04 server distro. Independently with the same hardware and network resources. I've monitored existing services on the machine, such as MySQL, PHP, and Apache2, but also coded dummy scripts to test individual features.

1. Platform functionality overview

✓ - partial or limited feature

✓✓ - feature is present

X - the platform does not offer this feature

Dynatrace

Dynatrace is an end-to-end observability platform offering tools mainly focused on monitoring modern infrastructures and mainly distributed applications, user experience, and business intelligence. Compared with Datadog, Dynatrace offers more advanced and better integrated AI-powered features, emphasizing APM.

Datadog

Datadog is praised for infrastructure and security monitoring features. Its APM solution offers end-to-end application performance monitoring enabling you to collect and monitor requests, traces, and logs and, with advanced features, also correlate this data and create actionable insights. Compared with Dynatrace, it handles the entire DevOps and SRE workflow, including almost complete Incident management and SIEM.

2. Ease of integration and deployment: Point Dynatrace

Deploying Dynatrace is fairly straightforward. The initial setup process offers sufficient onboarding support for deploying the agent based on the environment. The agent deployment required the execution of three code snippets provided by Dynatrace.

Maybe the most significant advantage of using Dynatrace is the fact that it allows you to configure the agent from the web UI. This made the setup of log monitoring but also APM relatively seamless.

After installing the OneAgent, Dynatrace automatically received and started reporting on Logs, Infrastructure metrics and proposed a list of used technology on my host. To start gathering distributed traces, I instrumented my application code with Open Telemetry. I especially appreciate that Dynatrace's documentation encouraged me to do so. To instrument the application, I followed the process described in Dynatrace's documentation to properly forward data to Dynatrace.

To start with Datadog, I firstly installed and configured the Datadog agent. This process was approximately the same Dynatrace. The first difference lies in configuration. Unlike Dynatrace, I had to visit multiple configuration files, starting with /etc/datadog-agent/datadog.yaml and then technology-specific config files like /conf.d/apache.d/conf.yaml . The reason for this was that I needed to enable log capturing.

To set up APM, I've decided to instrument the python dummy code with Datadog's python SDK dd-trace. While Open Telemetry and Tracing is available, it's not encouraged like in the case of Dynatrace, so I've decided to stick with vendor-provided guides. Datadog offers a web-UI configuration tool allowing you to create your own execution script, configuring additional parameters like environment, service name, traces sample rate, logs injection, or code profiling.

🔭 What is Open Telemetry?

OTel, for short, is a product of a merger of OpenCensus and OpenTracing in the interest of having one single standard. It's not an observability back-end like Datadog or Dynatrace. As defined in the documentation: “OTel's goal is to provide a set of standardized vendor-agnostic SDKs, APIs, and tools for ingesting, transforming, and sending data to an Observability back-end (i.e., open-source or commercial vendor)."

Datadog's initial setup is a bit more complex, with positive and negative implications. Dynatrace hoards data from all over the architecture practically automatically, which is great during the trial period, but it can also hurt your budget, in case you keep capturing data you don't actually use. In the case of Datadog, you have to do a lot of groundwork on your own, which enables a more tailored solution, but it also requires a lot of patience and also knowledge to match Dynatrace.

3. Onboarding, UI & UX: Point Dynatrace

I find Dynatrace's UI a bit overwhelming. While it's great that the deployment is practically automated, the sheer amount of data suddenly popping from all over the UI is hard to grasp initially. I also had to leave the UI multiple times to actually figure out the used terminology or tool grouping.

Dynatrace's documentation offers more than sufficient support to deploy, set up, and tweak the OneAgent. Anything from the general overview to instrumentation with open-telemetry was easy to find and follow. Dynatrace's university is available directly from the UI via a link in the user settings drop-down menu. The onboarding courses are in the form of modules, so short videos comparing Dynatrace to competition and explaining fundamental principles. While it's a great place to start, I think it's not a match for Datadog's learning center.

Datadog has a simple UI, which has some minor design flaws (like broken dark-UI in APM configuration) and some questionable grouping logic. But apart from that, it makes it easy to access data from multiple points based on the context and approach to said data. One can use either the Event explorer or individual product sub-pages to query and analyze ingested data.

Datadog's documentation is not very well-structured, but still well-written if one finds the right page. Compared with Dynatrace, Datadog's learning platform goes a step further and, apart from onboarding lectures-like videos, offers web-based coding labs, which actually enable new users to get hands-on experience in a simulated environment. I prefer this approach to onboard since it plunges users into the workflow from the start.

4. Incident management and Alerting: Point Datadog

Dynatrace offers minimal alerting and almost no incident management features out-of-the-box. You can access issues from the problems UI and set up monitoring based on events in the following categories: Monitoring unavailable, availability, error, slowdown, resource, and Custom.

🔮 Want to collaborate on solving incidents from one place?

Go to Better Uptime and start managing your incidents in 2 minutes.

Datadog offers an almost complete Incident Management tool. Everything except on-call schedules and incident communication (e.g., status pages) can be covered by Datadog. Datadog also makes it fairly easy to declare and manage incidents from events and monitors. With Datadog, you can create incidents, rank them by severity, manage incident resolution by assigning responsible users and teams, draft post-mortems, and send basic e-mail and slack notifications.

5. Pricing

Dynatrace offers both a full-stack pricing model, but also individual product pricing models. The Full-stack tier starts at $74 per 8GB per host. However, tools like Application Security work as add-ons and aren't included in the pricing. Dynatrace is available in two different deployment modes, either as a SaaS or managed Dynatrace, hosted on a cloud or private data center of your choice. The pricing is not unified in billing logic, ranging from prepaid Digital Experience Units to Dynatrace-specific Davis Data Units (DDUs). Luckily pricing FAQ explains both.

Digital Experience Units Depend on the number of synthetic monitors and user sessions. DDUs depend on the volume of external data ingested into the platform. So, e.g., amount of log lines and events or the number of custom metrics from cloud services.

Datadog has a decentralized pricing model. Each product has its own pricing logic. For example, APM & Continous profiler is priced per host. The same applies to Infrastructure monitoring, a requirement for APM. Other features, like log management, are billed based on ingested data, and incident management is billed per contributing user. Datadog has no full-platform pricing available. It's arranged with their sales team, with the option of volume discounts.

6. Third-party and community reviews: Tie

Apart from established third-party reviewing platforms like G2, Gartner, or TrustRadius, also community sites like Reddit or Twitter offer a general idea of what's it like to work with both tools.

Dynatrace takes pride in being the leader in Gartner's Magic Quadrant for APM and Observability. But clearly, as the scheme states, they come very close with Datadog. In terms of public opinion, Dynatrace generally does better in the APM ratings, mostly being praised for the automated setup, powerful AI tools, and complex approach to data. On the other hand, users tend to complain about alerting and false alarms, complex UI, and initial hardships in learning.

Datadog is praised for infrastructure and security monitoring and is often the first choice for those with these as an utmost priority. Conflicting reviews appear mainly in terms of platform usability with other stacks. Whereas one side brings up comprehensive support or integrations, others are afraid of potential vendor-lockin, with the requirements Datadog puts on some deployments (like APM and Infrastructure, for example).

Generally, in terms of community, Datadog has a wider footprint, allowing you to source many personal experiences of users on the internet, with many endorsements for both positive and negative experiences.

In summary, Datadog is a people's choice for infrastructure monitoring, security, and some business intelligence uses. Dynatrace is praised mainly in the field of APM.

Final Thoughts

Personally, I don't have a favorite this time. While I can clearly see the differences in both tools, their impact is tough to grasp. Datadog's complex pricing is prone to human error, like accidental host deployment, and New Relic's fully automated setup might have the same consequence with ingesting redundant data. While Datadog requires you to custom-tailor your deployment, it also lacks documentation, which makes it a bit more stressful.

To wrap things up, here are the main differences summed up.

If I were to deploy Dynatrace, I would certainly leverage the fully-automated data ingestion and compensate for the lack of Incident management with Better Uptime. In the case of Datadog, I would surely enhance the infrastructure monitoring with a log management solution like Logtail and potentially also plug in a more powerful APM solution like Sentry, New Relic, or other frontrunners from the Magic Quadrant.

If you want to explore more options please read our Dynatrace alternatives article.