Due to the Broadcom acquisition of CA Technologies, FreshTracks will be shutting down effective today. Thanks to everyone in the community for your support along the way! @PrometheusIO@CNCF@Grafana https://t.co/lHPAwbYydV
In Part 3 of his blog series, @limscoder explains how to use the open source data-sidecar project to execute computationally intensive multi-variate models against @PrometheusIO data.
https://t.co/322rppbgJI
Interested in executing more complex multivariate models with @PrometheusIO data? Check out part 2 of @limscoder’s latest blog series to explore TensorFlow Model Execution for time series data. https://t.co/qF7aHKvkGZ
Use Serverless functions (FaaS) for simpler deployment and easier scaling of #Prometheus rule evaluation. @limscoder walks through application creation techniques to handle computationally intensive analysis methods. https://t.co/qhncaFOdXg
Welcome to a new episode of "Things That Can Go Horribly Wrong With Your #Monitoring Infrastructure!" @Strofcon covers cardinality explosions in #prometheus and what you can do about them. https://t.co/1jIDtInEY7
@mchipouras talking about #Kubernetes#monitoring and observability at Comcast DevOps Day Denver 2018. You need both to look at predefined sets of metrics or logs and to explore properties and patterns that have not been defined in advance!
Dave Thompson, aka @limscoder, shares his expertise on load testing @PrometheusIO metric ingestion and the alternative implementations to meet the diverse use cases. https://t.co/bVwOdjqRr4
In case you missed it, this 6-part blog series by @bob_cotton does a deep dive into the sources of #Kubernetes metrics, how to collect them and how to alert on the important ones. https://t.co/Dx4jeTUVna
Wow! Very exciting day. It's been a long road to get here, but 'Cloud Native DevOps with Kubernetes' is now available for pre-order: https://t.co/2uQI3T99y6 (Amazon UK: https://t.co/w2PkiW9mfW) We're really happy with the cover, too! Please RT if you're interested in the book.
So, what do the #Kubernetes objects tell you? Using kube-state-metrics you can gather counts, labels, creation time and generic "info" to create interesting alerts and dashboards. Grab the snippets to get started.
Are you using kube-state-metrics to expose your cluster states? In part 6 of this blog series, @bob_cotton covers the cluster states you can gather and includes a list of the 19 #Kubernetes object types. https://t.co/wenfLIMbrj
10 nodes
158 containers
more than 273,000 unique series exposed from the cluster !!!
I recommend watching this to understand which metrics are exposed from which native and addon cluster components, and how to apply practical strategies to them to understand our apps.
To my fellow Chronomancers, let's talk about how we time align the output of our anomaly detection infrastructure with the upstream @PrometheusIO series at #PromCon2018. This lightning round is fast so bring your best time shifting skills.