Talk

People say: “You build it! and you own it!” but in the modern IT world, it is easy to write yamls and apply them to k8s, right? But it does not not always work as expected, doesn’t it?
You can find tons of books and tutorials on how to do it correctly. But the reality is that there are no manuals on how to write apps that are easier to debug and troubleshoot in k8s. How can the design and construction of k8s resources be improved to avoid bugs?
In this talk, I’ll show how to deal with the most common errors while deploying to k8s, different strategies and methods for troubleshooting and how to debug your apps inside the cluster.
I’ll show the most valuable tools you can use to speed up your debugging skills in RBACs, NetworkPolices and Nodes, volumes, and pods. Also, we discuss high-level strategies and methods for addressing diverse k8s failures and specific techniques to apply when writing, using, and running your projects inside k8s to make programs more accessible to debug.
Andrii Soldatenko
Dynatrace
About me: :sweat_smile:
My name is Andrii Soldatenko. I am a sr. software engineer originally from Ukraine :flag-ua:, and I am currently living in :flag-at: . Public speaker (KCD, FOSDEM, GoDays, PyCons) and OSS contributor (Apache Airflow, Golang, OpenAPI, docker). I am a big fan of debuggers, Neovim, Rust
 and I have a blog at http://asoldatenko.org. I am a CrossFit AMATEUR. Big fun of hand walking and different gymnastics 🙂
Python developer in the day, Go developer (gopher) under the hood. Big fan of full-text search and graph databases.
Contributed to different open source projects:
Apache Airflow (airflow helm chart and other related projects).
docker registry
pyhelm, aiohttp-swagger, mezzanine
chalice, requests, aiohttp tutorial
sendgrid-python and sendgrid-django
OpenAPI v3 specification, fix Go docs