Slight Reliability Slight Reliability Episode 46 - Raw Telemetry
In this episode I propose the use of scatterplots of raw data to better understand how our systems are behaviour and what our customers are experiencing. The ideas from this episode come from my time as a performance engineer and working with legends in that space Richard Leeke (https://www.linkedin.com/in/richard-leeke-450448/) and Neil Davies (https://www.linkedin.com/in/neildaviesnz/).
For some basic examples of scatterplots and what they show you versus line charts check out an article I wrote back in 2017 called "Let's Talk About Averages": https://www.linkedin.com/pulse/lets-talk-averages-stephen-townshend/
Another proponent of scatterplots is Stijn Schepers (https://www.linkedin.com/in/stijnschepers/). Here's an article he wrote about it in 2019: https://www.linkedin.com/pulse/performance-testing-act-like-detective-use-raw-data-stijn-schepers/
Neil Davies' article on tornado scatters "Chasing Tornadoes" can be found here: http://www.performance-workshop.org/wp/wp-content/uploads/2013/12/Chasing_Tornadoes_Davies.pdf
You can find the official Slight Reliability podcast website at: https://slightreliability.com/
You can find Stephen at:
LinkedIn: https://www.linkedin.com/in/stephentownshend/
Twitter: https://twitter.com/the_kiwi_sre