If you’re new here, read below for the 10,000 foot view of Datadog.
Your application stack is teeming with unused metadata that’s trying to tell a story: things aren’t humming along as well as they should be.
Exception thrown; Database queries slowing; Cache misses rising; Upstream service flapping; Error log growing. Each of these chatters a part of the story, but they’re hard to make sense of—or act on meaningfully—when taken separately from the others.
Datadog lets you collect all these metrics, events, and service states in one place. Then, visualize and correlate the data with beautiful graphing and set flexible alerting conditions on it—all without running any storage or monitoring infrastructure yourself.
Collect a wealth of already-available data without writing any code. Install the Datadog Agent everywhere—every server, instance, VM, node, container-running host—and then enable and configure any of our 200+ out-of-the-box integrations to start the metrics flowing to Datadog’s backend.
Submit custom application metrics by writing a little code. Instrument your own gauges, counters, timers, and histograms with DogStatsD, or use APM to trace the execution time of any code path to see how it impacts overall request-response times.
The client libraries for these send your custom metrics and trace data to the Datadog Agent, which ships them off to Datadog.
Some of your stack may be SaaS, not servers. Datadog can poll many of these services, and the integrations for them are the easiest of all to install; no Agent required.
As soon as you’re capturing all this data, see it immediately in the Datadog web application. Use the Metrics Explorer to search for a given metric and watch it ebb and flow. View and comment on events (say, an application deploy) as they pour into your Event Stream. Filter for some group of hosts in the Infrastructure Map. Get an overall picture of how some service (say, MySQL) is running via its default dashboard.
Before long, you will create custom Screenboards, where you will combine all the graphs, numbers, events, and service states you care about the most. You can customize the graphs in whatever way helps you suss out problems—skewing their metric values using other metrics, applying anomaly, outlier or forecasts detection, overlaying events onto them, and more.
Once your graphs have exposed problem areas, set some alerting conditions on your metrics using Monitors. You will get emails when the alerts fire, but set up the Slack or HipChat integration to get a dedicated notifications.
When you’re well aware of an ongoing problem, silence its alerts. When you’re about to bring a service down for maintenance, schedule a downtime so you won’t get spammed with alerts. When you can’t define some alert-worthy condition in terms of a single host, event, metric, or service, create a composite monitor.
You can contact our Support Team via email at firstname.lastname@example.org.
Chat with our support team any business day between the hours of 10:00 and 19:00 ET. You can reach us through in-app chat, or on Slack, Datadog Slack.
In-app Event Stream Messaging
Messages in the event stream containing @support-datadog will reach our Support Team. This is a convenient way to reference a graph snapshot.