Crispy Tech Tips

Cluster administrators usually need to collect Kubernetes events to allow a better troubleshooting. By default Kubernetes does not stream these logs on regular STDOUT, "k8s-event-logger" tools comes exactly for this reason!

Most Kubernetes pods could be better configured in terms of resources. Unfortunately, is not so easy out-of-the-box. Robusta KRR (Kubernetes Resource Recommender) comes to help us with it.

Optimize distributed builds with remote docker cache

Gone are the days when we had our applications always up&running. Cloud providers gave us a lot of elasticity, but without a mind change all this magic world will soon become a hell.

GKE autopilot is the Kubernetes solution by Google for who wants to run pods and nothing else while having in the meantime a Kubernetes feeling!

Speed up your container image builds with few changes on your Dockerfile

Standard time and daylight time have been a nightmare for cron configuration in Kubernetes. Fortunately it is an old nightmare thanks to the cronjob .spec.timeZone configuration

"With great power comes great responsibility" ... and infrastructure cost changes are part of that power

ReadWriteOnce, ReadOnlyMany, ReadWriteMany access modes don't make much sense IMO, new ReadWriteOncePod does

Who pays for shared resources? This is a gold question in every shared/multi-tenancy Kubernetes cluster.. Today I'll report a couple of crispy tools to deal with that.

easy proof of concept, presentation, education, hobby.. whatever the need is, you can use kodekloud's public playground to run for free a specific Kubernetes version.

Pod resource resizing is a work for professional craftsmen. Under estimation / overbooking are big enemies out there. Do you know that you can increase a pod resource without any container(s) restarts needed?

How many "docker tag .." have you done in your whole life?! (Don't lie) Crane allows you to easily move an image directly from the source to the destination in a single command way. That's great!

Until 26 of November 2023, the main way to have a managed mTLS solution on AWS was through API Gateway service. ... but a new feature is out there.. ALB is now supporting mTLS too!

Helm charts, helm charts everywhere! And what about changelog of helm charts?!? Unfortunately this is NOT an old nightmare :D but comparing different versions on artifacthub could make the nightmare less scary...

"I run k8s on a cloud provider so I don't care about worker nodes' logs and I'm so happy of that" Some days later.. Writing on google "How to access k8s worker node's logs in a SaaS solution ..."