The Sins of High Cardinality in Observability Metrics Collection
Offered By: USENIX via YouTube
Course Description
Overview
Explore strategies for managing high cardinality in observability metrics collection during this 20-minute conference talk from SREcon24 Americas. Delve into the challenges posed by granular labels and metadata in time-series storage and compute resources, which can lead to exponential growth and increased costs. Learn how to strike a balance between desired granularity and practical constraints in observability data. Discover various approaches to mitigate the impact of high metrics cardinality, with a focus on open source Prometheus metrics collection examples. Gain insights on optimizing your observability infrastructure to efficiently discover and remediate problems without compromising on essential granularity.
Syllabus
SREcon24 Americas - The Sins of High Cardinality
Taught by
USENIX
Related Courses
Site Reliability Engineering: Measuring and Managing ReliabilityGoogle Cloud via Coursera Introduction to DevOps and Site Reliability Engineering
Linux Foundation via edX Developing a Google SRE Culture
Google Cloud via Coursera Preparing for Google Cloud Certification: Cloud DevOps Engineer
Google Cloud via Coursera Organizational Change and Culture for Adopting Google Cloud
Google Cloud via Coursera