2 years ago

When it comes to metrics, cardinality is an important topic. For those who are not familiar with cardinality in metrics, it refers to the number of possible time series there can be, based on the dimensions the metrics have.

Chronosphere has written several articles on our blog related to https://chronosphere.io/learn/what-is-high-cardinality/, https://chronosphere.io/learn/classifying-types-of-metric-cardinality/ in your workload, https://chronosphere.io/learn/wrangle-metric-data-explosions-with-chronosphere-profiler/ and more. One thing that has not been covered well when it comes to cardinality is how important it is to understand cardinality over time.

The higher churn and resulting cardinality from cloud native workloads can easily overwhelm older time series databases, which were designed with fewer, longer-lived time series than we see today.