With observability costs skyrocketing due to the high cardinality of cloud-native environments, many companies are mandating cuts (also known as caps) to control spending. Within observability teams, everyone knows there is waste, but using traditional tools to identify and remove it is risky and takes time. A wrong decision can make troubleshooting harder and lead to lengthy incidents and outages.
This is a how-to session showcasing how you can analyze and shape your cloud-native observability data to reduce costs while fulfilling your existing dashboard and alerting needs. It allows you to control long-term data growth while ensuring your engineers have the necessary information to do their jobs.
In this session, you will see how to:
-Set Quotas by service, team, etc. to contain cardinality explosions and overages
-Analyze any metric to understand how much it is used, where, and by whom to weigh cost versus utility
-Apply aggregations to shape raw metric data volumes to improve performance and reduce costs
-Automatically optimize slow queries to ensure dashboards are performant