StatusNeo

Watchdog: Sniff Anomalies using autonomous Machine Learning

The more systems you have, the harder it becomes to watch over all of them. When your dynamic infrastructure includes tens of thousands of host containers and services, you can’t always anticipate where an issue will originate or what it will look like. That is why Datadog created Watchdog – an auto-detection engine that watches all your systems for signs of trouble.

Watchdog starts monitoring your applications automatically. You don’t need to train it or give it any commands. It is built on Datadog’s machine learning algorithms that continuously analyze data from across your systems. balsamiq wireframes license key free

Watchdog looks for irregularities in metrics, and pings for high and low signal/noise indicators. For eg. a sudden spike in the hit rate. For each irregularity, the Watchdog page displays a story. Each story includes a graph and summary of the issue in your feed and helps you get to the root cause faster. It also maps timeframes in series. To avoid false alarms, it reports issues after observing your data for a sufficient amount of time to establish a high degree of confidence. thx spatial audio activation code free

Story Details

The graph in the story shows the latency values of ELB different availability zones. Watchdog detected similar anomalies in this metric from a single load balancer enabled in three availability zones and automatically grouped these searches in a single story.

Expected Bounds
It tells about the upper and lower thresholds of expected behavior on the graph. Watchdog tracks key application metrics such as latency or error rate but it also watches other data to tell you when a networking issue in the cloud may be affecting.

Archiving Stories
You can archive that particular story from the feed, as well as from the home page. If a story is archived, the Yellow Watchdog binoculars icon is not displayed next to the relevant service.

Monitors
Monitors identified with your stories are displayed at the bottom. Each monitor showed up has the metric of the current story and its associated tags included in its scope.

“As the complexity of applications explodes, the need for automated issue detection becomes a necessity for teams to build highly performant and reliable applications in the cloud,” said Homin Lee, Head of Data Science at Datadog.