Applied online, HR quickly reached out to me for an initial discussion. They mentioned there would be fast interview sessions. There was no response for a week until I emailed them. They replied, unfortunately, there was a candidate with a stronger background.

I almost forgot about it until 2 weeks later when HR contacted me again, saying they would reconsider me to move on.

This was followed by a hiring manager interview. They were located in Greece and had a strong accent. I passed it with positive feedback, and then HR scheduled two technical interviews for me.

Round 1 Tech Review:

Two senior engineers with strong accents, located in the EU from different countries, asked some basic (not hard) questions related to Linux, Python (I told them my proficient language), and Kubernetes. Examples include:

• What does 'load average' mean? Is it busy or not if the load average is 2 when there are 100 CPU cores?
• What are the decorator keywords in Python?
• How do you back up the data of a statefulset resource in K8s?
• What happens behind K8s when you commit a kubectl create pod command?
• How do you limit K8s workloads to import images from a specific site?

Round 2 Tech:

I was interviewed by an SRE/DevOps architect. They asked me to explain a project I was most proud of, the difference between admission control and ingress, how to scale K8s workloads for pods and nodes, and to brainstorm considerations when maintaining/deploying more than 1000 K8s clusters.

Personally, I suppose I performed pretty nicely in the interviews. The only issue was that sometimes I could not understand their questions due to their strong accents.

Three days later, HR told me there were stronger candidates than me. Total waste of time.

Load average: what does it mean? Is it busy or not if the load average is 2 when there are 100 CPU cores?

What are the decorator keywords in Python?

How do you back up the data of a StatefulSet resource in K8s?

What happens behind K8s when you commit a kubectl create pod command?

How to limit K8s workloads to import images from a specific site?

Explain a project you are most proud of.

What is the difference between admission control and ingress?

How do you scale K8s workloads for pods and nodes?

Brainstorming: what to consider when you need to maintain/deploy more than 1000 K8s clusters?