Rushikesh Ujjainkar

[Rushikesh Ujjainkar @ujjainkar24]

@RajThackeray What about your party workers beating innocent people

[Rushikesh Ujjainkar @ujjainkar24]

@flamboypant Sahi Mai bhai

[Rushikesh Ujjainkar @ujjainkar24]

@anubhav__tweets Are bhai kya problem hai usme he don't want to involve in politics as simple as that

[Rushikesh Ujjainkar @ujjainkar24]

@IamVishal_17 @ThandaPeg Bhai send me as well

[Rushikesh Ujjainkar @ujjainkar24]

@Junethethree @litteralyme0 Bsk ab tak to unhi se kiya

[Rushikesh Ujjainkar @ujjainkar24]

@sauravstwt naukri

[Rushikesh Ujjainkar @ujjainkar24]

@RVCJ_FB 25% luck 75% hardwork 🔥

[Rushikesh Ujjainkar @ujjainkar24]

@RVCJ_FB Ganguly

[Rushikesh Ujjainkar @ujjainkar24]

@devops_nk C

[Rushikesh Ujjainkar @ujjainkar24]

@SajalDhiman2 Informative bro 💯

[Rushikesh Ujjainkar @ujjainkar24]

3️⃣ Still no heartbeat after 5 min (pod-eviction-timeout) → pods get evicted
Root causes I check first:
→ Disk full (DiskPressure)
→ kubelet/containerd crashed
→ Node → API server network broken
→ Clock skew / expired kubelet cert
→ Underlying EC2 instance impaired
#devops

[Rushikesh Ujjainkar @ujjainkar24]

What does "Node NotReady" actually mean?
It's not the node dying — it's the control plane not hearing from it.
Here's the chain:
1️⃣ kubelet sends heartbeats to API server every ~10s
2️⃣ No heartbeat within 40s (node-monitor-grace-period) → Ready flips to Unknown

[Rushikesh Ujjainkar @ujjainkar24]

@fromcodetocloud All the best for future bro 👍😊

[Rushikesh Ujjainkar @ujjainkar24]

@AskYoshik Thank you bro for this insightful.

[Rushikesh Ujjainkar @ujjainkar24]

@devops_nk Haha, hopefully soon bro . Working towards it and looking for the right opportunity. All the best for your next journey. ❤️

[Rushikesh Ujjainkar @ujjainkar24]

@uday_devops Claude

[Rushikesh Ujjainkar @ujjainkar24]

@cricketplusmem_ Morgan

[Rushikesh Ujjainkar @ujjainkar24]

@devops_nk Relatable bro 😢

[Rushikesh Ujjainkar @ujjainkar24]

5/ Check certificates
kubeadm certs check-expiration

6/ Verify critical pods rescheduled
kubectl get pods -o wide

Bonus: K8s auto adds NotReady taint
→ Pods evict after 5 mins
→ Reschedule on healthy nodes automatically
Save this for your next interview 🔥
#Kubernetes #DevOps

[Rushikesh Ujjainkar @ujjainkar24]

Got this in a K8s interview 👇

"Node is NotReady at 2AM,
3 critical pods are down.
What do you do?"

Here's how to answer it 🧵

1/ Check node status
kubectl get nodes
kubectl describe node <name>

2/ Check Conditions
❌ MemoryPressure → free -h

[Rushikesh Ujjainkar @ujjainkar24]

❌ DiskPressure → df -h + docker system prune
❌ PIDPressure → kill heavy processes

3/ SSH into node
systemctl status kubelet
systemctl restart kubelet
journalctl -u kubelet -n 100

4/ Check CNI plugin
kubectl get pods -n kube-system
→ Look for Calico/Flannel issues