Datadog vs Prometheus

When we are talking about network monitoring service, the list seems never ended and it keeps growing each years as the demand continue to increase. Among the big name you always heard, there are also some new players. In today Datadog vs Prometheus article, we are going to give you information about these two similar yet different services regarding what they can offer to you, so you can choose based on your preference. If one or both of them took your interest, go check our article below.

In this article, we are going to give you information about:
– What are Datadog and Prometheus
– How much is Datadog Subscription Cost
– What Datadog and Prometheus can offer to you
– Datadog vs Prometheus

About Datadog
Datadog is classified as infrastructure management and network management. The company offer their product as software as service and to use them, you have to install an agents that you can use in various platforms. Its capabilities are including high-end analysis features and customization through a variety of application programming interfaces. Their service is specially designed for hybrid cloud environments and it able to monitor the performance of network, apps, tools, and services as well as compatible with many different popular alert and teaming service like Atlassian and Slack. Read also: Datadog vs Splunk here.

Datadog Pricing
To use the service, you can start from free account, which is completely free as long as you want but of course with limitations like only up to 5 host and 1-day metric retention. As for the paid version, the price will start at $15 per host per month or the most expensive plan that start at $23 per host per month. These paid subscription will give you more than 250 integrations, out of the box dashboards, machine learning-based alerts and live processes.

Datadog Capabilities
To use Datadog, you will have to install and configuring the software agents first, which available for a host of platforms like OS X, Chef, CoreOS, Debian, Docker, Windows 10, Puppet, Fedora, and Ubuntu. After the agent is installed, you have to connect it with your account to begin monitoring. You will also need set up the monitoring targets like Apache, Microsoft IIS and SQL or others and even though the level of efforts involved in setting them up are depend on the service, most of them will simply need configuring connection and credential information.

The service is able to monitor and give you alerts on several different source categories with different level of details in each one of them, so depend on what is being monitored, it can incorporate log files, availability, performance ,metrics, configuration and change tracking as well as other numerous information. There is a new log management in format of single pane of glass to monitor all of your logs, so you can move from a log to line to trace and to metrics easily.

Since Datadog focus is in real-time monitoring, it doesn’t offer a conventional reporting tools. However, they do offer analysis capabilities, which can probably surpassing traditional report set including the search capability in the Event view. You may also build queries based on a number of criteria like event priority, tag, the host, and status. Another strong point of the service is its level of integration and customization that let you have over 200 built in integrations with open source software, cloud services, enterprise apps, and standard-based admin tools.

About Prometheus
Another alternative you can pick if you don’t want to use a paid service is Prometheus. This one is an open-source systems monitoring and alerting that was initially built in SoundCloud. It becomes more popular and used by many companies or organization while having an active community of both users and developers. The project list themselves in Cloud Native Computing Foundation in 2016 as the second hosted project after Kubernetes to clarify that it is an open source project and maintained independently by any company.

Prometheus Capabilities
To start using Prometheus for monitoring your services, first you have to expose those services to a Prometheus endpoint, which is a HTTP interface that expose a list of metrics and the current value of the metrics. It will start retrieving data from your services with its wide range of service discovery options. Then, your services metrics interface will be polled by their server and your data will also be stored. You may write queries in PromQL language in its UI to extract metric information.

The endpoints we mentioned earlier will define the metrics and values for the metrics as well as exposed over HTTP. Prometheus metrics follow many guidelines set by metrics 2.0, which means the metrics have name, dimension, description, and values while the only missing out here is the unit for the metrics because these endpoints provide a standardized way of gathering metrics. There are many services exposing Prometheus endpoints, so gathering metrics from them are easy but in case they don’t, you will need a convertor.

For metrics, Prometheus will poll services, so just like we said earlier, all of your services should expose to its endpoint. It is integrated with Kubernetes and also uses service discovery to find all of your services, so once it found all services, it will gather metrics by polling their metrics endpoint. The best thing about this method is you don’t have to install an agent while the metrics can be pulled by multiple Prometheus instances.

Besides all the good points, Prometheus also have some downside. First is it is not made to solve several aspects, second is it don’t offer support for log but there are already many other log aggregators out there that you can use for free. It also does not offer dashboarding solution since it still rely on Grafana for dashboarding, which makes the usage becomes even more complicated.

Now, let’s compare Datadog with Prometheus. As you already know, Prometheus is open source, so it is free to use but of course in ease of use, it is more complicated than Datadog since you have to do a lot of things first before you can start monitoring your services. Another difference is Datadog is a complete solution while the other is only for gathering metrics.

Datadog vs Prometheus

- A complete solution- Doesn’t support logs
- Free & paid subscription- Free
- Good dashboarding- Not a dashboarding solution

All in all, the decision is all yours to make. In our opinion, both of them are made for different type of users, if your scale is smaller or are okay to use other log aggregator, Prometheus is a great option but if you want a more complete solution, go for Datadog.

Leave a Reply

Your email address will not be published. Required fields are marked *