The Datadog Agent is a piece of software that runs on your hosts. Its job is to faithfully collect events and metrics and bring them to Datadog on your behalf so that you can do something useful with your monitoring and performance data.
For information on running the Agent through a proxy, see the dedicated documentation Datadog Agent and proxy ; for which IP ranges to allow, see the complete list of IP’s and ports.
The Agent has three main parts: the collector, DogStatsD, and the forwarder:
The collector: runs checks on the current machine for whatever integrations you have and it captures system metrics such as memory and CPU.
DogStatsD: It is a statsd backend server you can send custom metrics to from an application.
The forwarder: retrieves data from both DogStatsD and the collector and then queues it up to be sent to Datadog.
This is all controlled by one supervisor process. We keep this separate so you don’t have to have the overhead of each application if you don’t want to run all parts, although we generally recommend you do.
Manage the Datadog agent and integrations using configuration management tools:
Chef, Puppet, and Ansible integrations using our public APIs so if you’re interested in using another automation tool, the above could be leveraged as examples to get you started.
There is also community support for Saltstack:
If you ended up at this page and have not yet installed the Datadog Agent, go to the dedicated agent integration page for installation instructions. If you just installed the Agent, it might take a few moments before you start seeing metrics appear. The first place you should check for metrics is the Metrics Explorer.
If you think you might be experiencing issues, the first thing to do is run the info command and check the Agent logs. The info command and the log locations are dependent on your OS, which you can select from the navigation to the left for further information.
If you encountered an issue during the Agent installation process that prevented installation from occurring, reach out to Datadog support team. Let us know your OS and version, as well as how you are installing the Agent. You should include any error messages you encountered along the way.
If you get the Agent installed but are not seeing any data in Datadog, you can troubleshoot in the following manner. First, run the info command. Select your OS in the navigation column on the left of this page to see how to run this. Does running the info command show any errors?
If not, you should also check the logs (location of the logs again depends on OS). Errors in the logs may also reveal the cause of any issues.
We have also seen a few cases where machines have their clock set further in the future or the past, which can sometimes cause problems with metric submission. To check for this, run:
date -u && curl -s -v https://app.datadoghq.com 2>&1 | grep Date
This outputs the current system’s date, and then makes a request to our endpoint and grabs the date on our end. If these are more than a few minutes apart, you should correct the time settings on your server.
Here is a quick guide for troubleshooting integrations installation:
No, it’s not:
Yes, it’s there:
system.cpu.userfrom the host that is running the Agent and has that integration setup.