Where we discuss scripted installs vs golden images vs immutable infrastructure via containers.

Where we discuss repeatability and reliability in systems design, why it’s important and what it means for your job.

Where we discuss pratical guidelines for expanding your monitoring platform to cover the new and challenging world of Microservice (and Service Oriented Architecture) monitoring, with brief side trips into log tracing and fostering healthy culture.

Where we discuss how to run statsd and metrics pipelines at a scale of over 750,000 packets per second.

Where we discuss the evolution of modern filesystems.

Where we discuss using Elasticsearch (along with Kibana, Timelion and Kafka) as a Timeseries Database (TSDB). It has several drawbacks over traditional timeseries storage engines, mostly in storage efficiency, but has other unique attributes that may overcome these drawbacks in the right use case.

Where we discuss histograms as the best data type ever, and the complexities of actually using them in Prometheus. For the most part, this is Jack Neely’s domain of expertise, so he does most of the talking in this episode.

Where we discuss how to find privacy on the internet, both for yourself at home and for users of your services.

Where we discuss the AWS S3 outage on February the 28th, which was due to an operations error, and relevant discussions about outages and how to learn from mistakes.