Your magic rancher cluster, your battle horse, the heart of the company, super secure, behind Single-Sign-On, it all seems beautiful, and then your SSO changes…

Imagine the scenario, your Kubernetes workloads have precious logs you wanna inspect and then someone does a rollout and all your pods are recreated… AAAAAAND it’s gone…

Because ruby in this case: just doesn’t cut it. And (ab-)using at_exit is not an option.

Finally understood how to make a helm deployment use another chart and configure it! I’ll show you how

The tale of how we found a very peculiar condition on ruby involving all the greatest hits of software: Threads, Processes, & C Compilers.

Or how to understand this sparse in-disk pointer table?

From the previous adventures in: Why my Kubernetes rails app returns 502 errors?, do I really need a Kubernetes cluster to experience an OOM in my workload? That seems excessive…

Why one app can handle >10k requests per second without sweating a bit [pun intended] and another one with two orders of magnitude less traffic gets sporadic 50X errors served by AWS ELB.

How many times you wanted to have your endpoint exclusively receive requests in JSON format?

old man yells at cloud.gif

It’s been a journey since Multi-factor authentication on SUSE’s Build Service @ suse.com blog, but now there’s code to show! 🎉

IT TOOK ME THREE MONTHS TO FIGURE THIS SHIT OUT.

Join me in this review of my day debugging an asset pipeline rewritten from Sprockets to Propshaft for the 3rd time.

Took me long enough to write some code & styles 😅, it should be bearable now.

TIL: distribution/distribution doesn’t allow foreign layers by default. You must provide an allow/deny list in the configuration.

Hey hey, I should pick this up and write down stuff :)