🗣️ 'I notice a lot of 5xx errors from the application during scale-in of the pods after a spike period.'

I was asked this question last week and spent some time debugging the issue. In the process, I realized there is confusion about how Kubernetes does (not do) graceful

And the solution for this week's challenge is… 🥁🥁🥁

Kubernetes will schedule 1000 pods on the same node, but not all will be running (answer: none of the above). (option 4).

You can find the full solution here: medium.com/@danielepolenc…

In the meantime, if you liked this

🏆 Kubernetes challenge: all you can eat.

I'm authoring questions for our Kubernetes courses at Learnk8s and wanted to share them with you! This week is about requests and scheduling!

You can find last week's challenge on rolling updates here medium.com/@danielepolenc…

---

In

And the solution for this week's challenge is… 🥁🥁🥁

There are no default probes. (option 1).

You can find the full solution here: medium.com/@danielepolenc…

In the meantime, if you liked this challenge, you might also like:

- The Kubernetes courses that we run at Learnk8s (we

Become an expert in Kubernetes!

Next week Learnk8s runs a 4-day Advanced Kubernetes course!

If you want to get your hands dirty with Kubernetes, join us for a workshop packed with hands-on labs!

Sign up here: learnk8s.io/online-advance…

Become an expert in Kubernetes!

Learnk8s runs a 4-day Advanced Kubernetes course on the 18th of April!

If you want to get your hands dirty with Kubernetes, join us for a workshop packed with hands-on labs!

Sign up here: learnk8s.io/online-advance…

This week on the Learn Kubernetes Weekly:

👀 How we preview Kubernetes changes
🎬 ArgoCD finalizer
🥷 Npm packages caught exfiltrating Kubernetes Config
🎫 Managed SSL certs for a private cluster
📊 Kubernetes Resource Recommender

Read it now: learnk8s.io/issues/73

What's better, one large Kubernetes cluster or many small ones?

Hear what Daniele Polencic — @[email protected] (@learnk8s ) & Michael Peterson (@loft_sh ) have to say about it -youtube.com/watch?v=LIITGz…

🏆 Challenge: Rollin'.

I'm authoring questions for our Kubernetes courses at Learnk8s and wanted to share them with you! This week is about rolling updates!

You can find last week's challenge on the readiness probe here medium.com/@danielepolenc…

---

In Kubernetes, you often

And the solution for this week's challenge is… 🥁🥁🥁

Answer 2!

Applications deployed on Kubernetes should use the readiness probe only for themselves — you should not check dependencies.

You can find the full solution here: medium.com/@danielepolenc…

In the meantime, if you

Become an expert in Kubernetes!

Learnk8s runs a 4-day Advanced Kubernetes course on the 18th of April!

If you want to get your hands dirty with Kubernetes, join us for a workshop packed with hands-on labs!

Sign up here: learnk8s.io/online-advance…

This week on the Learn Kubernetes Weekly:

🏅 Golden testing Helm
📈 Harnessing the power of metrics
1️⃣ Kubernetes Gateway API v1.0
📽️ How to use TokenRequest API and TokenVolume Projection
🪨 ReadOnlyRootFilesystem

Read it now: learnk8s.io/issues/72

🏆 Challenge: Readiness probes.

I'm authoring questions for our Kubernetes courses at Learnk8s and wanted to share them with you! This week is about rolling updates!

You can find last week's challenge on HA control planes here medium.com/@danielepolenc…

---

In Kubernetes, you

And the solution for this week's challenge is… 🥁🥁🥁

Answer 2 or 3

Only one scheduler and one control plane can be active at any point in time.

So you can't end up with three controller managers creating 3 ReplicaSets and 9 pods.

You can find the full solution here:

This week on the Learn Kubernetes Weekly:

💨 Airflow on Kubernetes for 2 years
📝 Learning apple/pkl
👋 Migrating from Pod Security Policies
👷🏻‍♂️ Build a Lightweight Internal Developer Platform with Argo CD and Kubernetes Labels

Read it now: learnk8s.io/issues/71

Having multiple tenants sharing a Kubernetes cluster makes sense from a cost perspective, but what’s the overhead? Read more about this in this latest blog post from Daniele Polencic — @[email protected] ⭐

eu1.hubs.ly/H087nQj0

🏆 Challenge: Multiplying pods.

I'm authoring questions for our Kubernetes courses at Learnk8s and wanted to share them with you! This week is about rolling updates!

You can find last week's challenge on preStop hooks here medium.com/@danielepolenc…

---

In Kubernetes, you can

And the solution for this week's challenge is… 🥁🥁🥁

Option 3 'The cluster is slower'!

This is counter intutive at first, but the more etcd nodes you have in your cluster, the slower it is to agree on a value written to disk.

So any operation (e.g. kubectl apply) take longer

How to choose the best node for your Kubernetes cluster before writing any code: lin0.de/QTsybx via Daniele Polencic — @[email protected]