AnveshGummala/k8s-troubleshooting-customdsv2

messages listlengths 3 3
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl describe pod shows ImagePullBackOff, what should I check?" }, { "role": "assistant", "content": "Check three things in order: (1) `kubectl describe pod <n...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "My pod keeps CrashLoopBackOff but I can't see why." }, { "role": "assistant", "content": "Start with `kubectl logs <pod> --previous` -- the current container may ...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A pod is stuck in Pending and never schedules." }, { "role": "assistant", "content": "Run `kubectl describe pod <pod>` and read the Events section -- the schedule...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl get pods shows a pod as Evicted." }, { "role": "assistant", "content": "Evicted almost always means the node ran out of a resource -- usually disk (`kubec...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "OOMKilled -- how do I actually fix this, not just restart it?" }, { "role": "assistant", "content": "Exit code 137 means the container hit its memory limit, not...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl rollout status deployment/myapp just hangs, what's happening?" }, { "role": "assistant", "content": "This usually means new pods aren't reaching Ready. Ch...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "How do I roll back a bad deployment quickly?" }, { "role": "assistant", "content": "`kubectl rollout undo deployment/myapp` reverts to the previous ReplicaSet. To...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A pod can't resolve other services by name -- nslookup fails inside the pod." }, { "role": "assistant", "content": "Check CoreDNS is actually running: `kubectl ge...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Ingress returns 502 Bad Gateway even though the pod is Running." }, { "role": "assistant", "content": "502 means the ingress controller reached a backend but got ...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl get svc shows my Service but curl to the ClusterIP times out." }, { "role": "assistant", "content": "Check the Service's selector actually matches labels ...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "PersistentVolumeClaim is stuck in Pending." }, { "role": "assistant", "content": "Check `kubectl describe pvc <name>` for events. Common causes: no StorageClass m...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "I deleted a PVC but the underlying disk is still there in my cloud provider." }, { "role": "assistant", "content": "That's expected if the StorageClass's `reclaim...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl gives me 'Error from server (Forbidden)' -- how do I debug RBAC?" }, { "role": "assistant", "content": "`kubectl auth can-i <verb> <resource> --as=<user-o...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A pod's application gets 403 errors calling the Kubernetes API from inside the cluster." }, { "role": "assistant", "content": "The pod is likely using its default...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A node shows NotReady in kubectl get nodes." }, { "role": "assistant", "content": "SSH to the node if possible and check `systemctl status kubelet` -- a crashed o...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "How do I safely take a node out of rotation for maintenance?" }, { "role": "assistant", "content": "`kubectl cordon <node>` marks it unschedulable without touchin...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "HorizontalPodAutoscaler isn't scaling up even though CPU usage looks high." }, { "role": "assistant", "content": "Check `kubectl describe hpa <name>` -- it shows ...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Cluster autoscaler isn't adding nodes even though pods are stuck Pending." }, { "role": "assistant", "content": "Check `kubectl describe pod <pod>` for the exact ...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "What's actually the difference between requests and limits?" }, { "role": "assistant", "content": "`requests` is what the scheduler uses to decide which node has ...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Pods fail to create with 'exceeded quota' errors." }, { "role": "assistant", "content": "Check `kubectl describe resourcequota -n <namespace>` to see the namespac...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "I updated a ConfigMap but my running pods still have the old values." }, { "role": "assistant", "content": "If the ConfigMap is mounted as a volume, updates do ev...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A Kubernetes Job shows 0/1 completions and never finishes." }, { "role": "assistant", "content": "Check the Job's pod directly with `kubectl logs job/<name>` -- i...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "docker pull against our internal registry fails with 'manifest not found' even though the tag exists." }, { "role": "assistant", "content": "This is a classic reg...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "docker buildx build --cache-to type=registry fails with 401 Unauthorized, but a plain docker pull of the same image works fine." }, { "role": "assistant", "conten...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "After a GKE node restarts, nvidia-smi fails with 'couldn't communicate with the NVIDIA driver' even though it worked before." }, { "role": "assistant", "content":...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "How does GKE decide whether to scale up a Spot node pool or a standard one?" }, { "role": "assistant", "content": "By default, GKE's cluster autoscaler is cost-aw...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "My SSH session to a cloud VM keeps dying and killing my long-running training job." }, { "role": "assistant", "content": "Run long jobs inside `tmux` (or `screen`...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "helm upgrade fails with 'another operation is in progress'." }, { "role": "assistant", "content": "A previous install/upgrade/rollback didn't finish cleanly, leav...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A StatefulSet pod is stuck Terminating for several minutes." }, { "role": "assistant", "content": "StatefulSet pods respect `terminationGracePeriodSeconds` (defau...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "My pod's liveness probe keeps restarting a container that looks healthy in the logs." }, { "role": "assistant", "content": "Check `kubectl describe pod <pod>` for...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "I added a startupProbe and now kubectl rollout status takes way longer than before." }, { "role": "assistant", "content": "That's expected -- `startupProbe` block...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A pod is stuck with STATUS Init:0/1 and never starts the main container." }, { "role": "assistant", "content": "Init containers run to completion sequentially bef...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "One container in my pod keeps CrashLoopBackOff but the main app container is fine." }, { "role": "assistant", "content": "Check that specific container's logs wit...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Pods in my namespace suddenly can't talk to each other after someone applied a NetworkPolicy." }, { "role": "assistant", "content": "Any NetworkPolicy that select...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "My pod is running but there's no istio-proxy container and traffic isn't going through the mesh." }, { "role": "assistant", "content": "Check the namespace label ...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A newly joined node shows NotReady and pods scheduled to it are stuck ContainerCreating." }, { "role": "assistant", "content": "This is almost always the CNI plug...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Some pods can reach a Service intermittently but not consistently, even though endpoints look correct." }, { "role": "assistant", "content": "Check `kubectl logs ...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "DNS resolution broke cluster-wide right after someone edited the CoreDNS ConfigMap." }, { "role": "assistant", "content": "`kubectl logs -n kube-system -l k8s-app...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "After a NetworkPolicy was added, pods can't reach an external API anymore, only internal services." }, { "role": "assistant", "content": "Any NetworkPolicy with a...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A pod can resolve a Service in its own namespace but not one in a different namespace by the same short name." }, { "role": "assistant", "content": "Short names l...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A NodePort service works with kubectl port-forward but not when hitting the node's external IP directly." }, { "role": "assistant", "content": "This is almost alw...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A Service of type LoadBalancer has stayed <pending> for the external IP for a long time." }, { "role": "assistant", "content": "Check the cloud controller manager...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl apply fails with an opaque error like 'admission webhook denied the request'." }, { "role": "assistant", "content": "The error usually names the webhook -...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl apply hangs and eventually fails with 'failed calling webhook: context deadline exceeded'." }, { "role": "assistant", "content": "This means the webhook's...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "I applied a CustomResourceDefinition but creating instances of it fails as if the type doesn't exist." }, { "role": "assistant", "content": "Check `kubectl get cr...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A custom operator's CR shows the same status forever and never progresses." }, { "role": "assistant", "content": "Check the operator controller's own pod logs, no...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A custom resource is stuck Terminating and won't actually delete." }, { "role": "assistant", "content": "A finalizer on the resource is waiting for something (usu...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Pods fail to create in a namespace with an error mentioning 'violates PodSecurity'." }, { "role": "assistant", "content": "Check the namespace's PodSecurity label...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "The whole cluster feels sluggish -- kubectl commands and pod scheduling are slow." }, { "role": "assistant", "content": "Check etcd health first, since it backs e...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl commands are timing out with 'connection refused' against the API server." }, { "role": "assistant", "content": "Check whether the apiserver's static pod ...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubeadm join fails with 'token has expired or is not valid'." }, { "role": "assistant", "content": "Bootstrap tokens expire after 24 hours by default -- generate ...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Upgraded the control plane and now some nodes show odd errors about unsupported API versions." }, { "role": "assistant", "content": "Kubernetes only supports kube...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Pods won't schedule onto a specific node pool even though it has capacity." }, { "role": "assistant", "content": "Check `kubectl describe node <node>` for a Taint...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Lower-priority pods keep getting evicted even though nothing looks wrong with them." }, { "role": "assistant", "content": "Check `kubectl get events --field-selec...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl drain hangs and never finishes evicting pods." }, { "role": "assistant", "content": "Check `kubectl get pdb -n <namespace>` -- drain respects PodDisruptio...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Pods stay Pending with 'did not match pod topology spread constraints' in the events." }, { "role": "assistant", "content": "Your `topologySpreadConstraints` with...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A Deployment can't scale past a certain number of replicas even though there's plenty of node capacity." }, { "role": "assistant", "content": "Check for `podAntiA...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "I have both a VerticalPodAutoscaler and HorizontalPodAutoscaler on the same deployment and scaling looks erratic." }, { "role": "assistant", "content": "VPA and H...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "HorizontalPodAutoscaler with a custom Prometheus metric shows <unknown> for the current value." }, { "role": "assistant", "content": "Query the custom metrics API...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Cluster autoscaler won't scale down a node even though it looks nearly empty." }, { "role": "assistant", "content": "Check for pods on that node with the `cluster...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A pod is stuck ContainerCreating with 'FailedAttachVolume' in events." }, { "role": "assistant", "content": "Check `kubectl describe pod <pod>` for the full event...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "PVCs across the whole cluster are stuck Pending and describe just says 'waiting for a volume to be created'." }, { "role": "assistant", "content": "Check the CSI ...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "I rotated a Secret's value but the pod using it as a volume mount still has the old one." }, { "role": "assistant", "content": "Same behavior as ConfigMaps: Secre...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "An app can't read its mounted config file even though the ConfigMap volume mount looks correct." }, { "role": "assistant", "content": "Check the volume's `default...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A pod shows ErrImageNeverPull even though the image is definitely present." }, { "role": "assistant", "content": "Check `imagePullPolicy` in the pod spec -- `Neve...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "My app gets intermittent 401s calling the Kubernetes API, but only after it's been running for a while." }, { "role": "assistant", "content": "Bound ServiceAccoun...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Pods fail to start with 'failed to create containerd task' in kubelet logs." }, { "role": "assistant", "content": "This is a container runtime-level failure, one ...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A node went NotReady with no workload changes, and kubelet logs mention a certificate error." }, { "role": "assistant", "content": "kubelet's client certificate r...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "After migrating nodes to a newer OS image, containers are getting OOMKilled at lower memory usage than before." }, { "role": "assistant", "content": "Newer OS ima...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A node hit DiskPressure and started evicting pods, but there's no obviously huge workload on it." }, { "role": "assistant", "content": "Unused container images an...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl top pod returns an error instead of numbers." }, { "role": "assistant", "content": "Check `kubectl get apiservices v1beta1.metrics.k8s.io` -- if it shows ...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl exec fails with 'error: unable to upgrade connection: container not found'." }, { "role": "assistant", "content": "This usually means the container name i...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl apply fails with 'the object has been modified; please apply your changes to the latest version and try again'." }, { "role": "assistant", "content": "Som...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "I applied a Kustomize overlay but the deployed resources don't have the patches I expected." }, { "role": "assistant", "content": "Render it locally first with `k...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "An ArgoCD Application keeps showing OutOfSync and re-syncing without ever settling." }, { "role": "assistant", "content": "Check `kubectl get application <name> -...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "helm install fails partway through and leaves resources behind." }, { "role": "assistant", "content": "Check `kubectl get pods -n <namespace> -l 'helm.sh/hook'` f...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Ingress TLS isn't working and the cert-manager Certificate resource shows Ready: False." }, { "role": "assistant", "content": "`kubectl describe certificate <name...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A Job's pod completed successfully but the Job itself shows Failed." }, { "role": "assistant", "content": "Check `kubectl describe job <name>` for `backoffLimit` ...
[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl delete namespace has been stuck in Terminating for a long time." }, { "role": "assistant", "content": "Almost always a leftover finalizer on some resource...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl describe pod shows ImagePullBackOff, what should I check?" }, { "role": "assistant", "content": "Check three things in order: (1) `kubectl describe pod <n...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "My pod keeps CrashLoopBackOff but I can't see why." }, { "role": "assistant", "content": "Start with `kubectl logs <pod> --previous` -- the current container may ...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A pod is stuck in Pending and never schedules." }, { "role": "assistant", "content": "Run `kubectl describe pod <pod>` and read the Events section -- the schedule...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl get pods shows a pod as Evicted." }, { "role": "assistant", "content": "Evicted almost always means the node ran out of a resource -- usually disk (`kubec...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "OOMKilled -- how do I actually fix this, not just restart it?" }, { "role": "assistant", "content": "Exit code 137 means the container hit its memory *limit*, not...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl rollout status deployment/myapp just hangs, what's happening?" }, { "role": "assistant", "content": "This usually means new pods aren't reaching Ready. Ch...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "How do I roll back a bad deployment quickly?" }, { "role": "assistant", "content": "`kubectl rollout undo deployment/myapp` reverts to the previous ReplicaSet. To...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A pod can't resolve other services by name -- nslookup fails inside the pod." }, { "role": "assistant", "content": "Check CoreDNS is actually running: `kubectl ge...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Ingress returns 502 Bad Gateway even though the pod is Running." }, { "role": "assistant", "content": "502 means the ingress controller reached a backend but got ...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl get svc shows my Service but curl to the ClusterIP times out." }, { "role": "assistant", "content": "Check the Service's selector actually matches labels ...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "PersistentVolumeClaim is stuck in Pending." }, { "role": "assistant", "content": "Check `kubectl describe pvc <name>` for events. Common causes: no StorageClass m...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "I deleted a PVC but the underlying disk is still there in my cloud provider." }, { "role": "assistant", "content": "That's expected if the StorageClass's `reclaim...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl gives me 'Error from server (Forbidden)' -- how do I debug RBAC?" }, { "role": "assistant", "content": "`kubectl auth can-i <verb> <resource> --as=<user-o...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A pod's application gets 403 errors calling the Kubernetes API from inside the cluster." }, { "role": "assistant", "content": "The pod is likely using its default...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A node shows NotReady in kubectl get nodes." }, { "role": "assistant", "content": "SSH to the node if possible and check `systemctl status kubelet` -- a crashed o...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "How do I safely take a node out of rotation for maintenance?" }, { "role": "assistant", "content": "`kubectl cordon <node>` marks it unschedulable without touchin...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "HorizontalPodAutoscaler isn't scaling up even though CPU usage looks high." }, { "role": "assistant", "content": "Check `kubectl describe hpa <name>` -- it shows ...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Cluster autoscaler isn't adding nodes even though pods are stuck Pending." }, { "role": "assistant", "content": "Check `kubectl describe pod <pod>` for the exact ...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "What's actually the difference between requests and limits?" }, { "role": "assistant", "content": "`requests` is what the scheduler uses to decide which node has ...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Pods fail to create with 'exceeded quota' errors." }, { "role": "assistant", "content": "Check `kubectl describe resourcequota -n <namespace>` to see the namespac...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "I updated a ConfigMap but my running pods still have the old values." }, { "role": "assistant", "content": "If the ConfigMap is mounted as a volume, updates do ev...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A Kubernetes Job shows 0/1 completions and never finishes." }, { "role": "assistant", "content": "Check the Job's pod directly with `kubectl logs job/<name>` -- i...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "docker pull against our internal registry fails with 'manifest not found' even though the tag exists." }, { "role": "assistant", "content": "This is a classic reg...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "docker buildx build --cache-to type=registry fails with 401 Unauthorized, but a plain docker pull of the same image works fine." }, { "role": "assistant", "conten...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "After a GKE node restarts, nvidia-smi fails with 'couldn't communicate with the NVIDIA driver' even though it worked before." }, { "role": "assistant", "content":...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "How does GKE decide whether to scale up a Spot node pool or a standard one?" }, { "role": "assistant", "content": "By default, GKE's cluster autoscaler is cost-aw...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "My SSH session to a cloud VM keeps dying and killing my long-running training job." }, { "role": "assistant", "content": "Run long jobs inside `tmux` (or `screen`...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "helm upgrade fails with 'another operation is in progress'." }, { "role": "assistant", "content": "A previous install/upgrade/rollback didn't finish cleanly, leav...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A StatefulSet pod is stuck Terminating for several minutes." }, { "role": "assistant", "content": "StatefulSet pods respect `terminationGracePeriodSeconds` (defau...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "My pod's liveness probe keeps restarting a container that looks healthy in the logs." }, { "role": "assistant", "content": "Check `kubectl describe pod <pod>` for...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "I added a startupProbe and now kubectl rollout status takes way longer than before." }, { "role": "assistant", "content": "That's expected -- `startupProbe` block...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A pod is stuck with STATUS Init:0/1 and never starts the main container." }, { "role": "assistant", "content": "Init containers run to completion sequentially bef...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "One container in my pod keeps CrashLoopBackOff but the main app container is fine." }, { "role": "assistant", "content": "Check that specific container's logs wit...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Pods in my namespace suddenly can't talk to each other after someone applied a NetworkPolicy." }, { "role": "assistant", "content": "Any NetworkPolicy that select...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "My pod is running but there's no istio-proxy container and traffic isn't going through the mesh." }, { "role": "assistant", "content": "Check the namespace label ...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A newly joined node shows NotReady and pods scheduled to it are stuck ContainerCreating." }, { "role": "assistant", "content": "This is almost always the CNI plug...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Some pods can reach a Service intermittently but not consistently, even though endpoints look correct." }, { "role": "assistant", "content": "Check `kubectl logs ...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "DNS resolution broke cluster-wide right after someone edited the CoreDNS ConfigMap." }, { "role": "assistant", "content": "`kubectl logs -n kube-system -l k8s-app...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "After a NetworkPolicy was added, pods can't reach an external API anymore, only internal services." }, { "role": "assistant", "content": "Any NetworkPolicy with a...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A pod can resolve a Service in its own namespace but not one in a different namespace by the same short name." }, { "role": "assistant", "content": "Short names l...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A NodePort service works with kubectl port-forward but not when hitting the node's external IP directly." }, { "role": "assistant", "content": "This is almost alw...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A Service of type LoadBalancer has stayed <pending> for the external IP for a long time." }, { "role": "assistant", "content": "Check the cloud controller manager...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl apply fails with an opaque error like 'admission webhook denied the request'." }, { "role": "assistant", "content": "The error usually names the webhook -...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl apply hangs and eventually fails with 'failed calling webhook: context deadline exceeded'." }, { "role": "assistant", "content": "This means the webhook's...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "I applied a CustomResourceDefinition but creating instances of it fails as if the type doesn't exist." }, { "role": "assistant", "content": "Check `kubectl get cr...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A custom operator's CR shows the same status forever and never progresses." }, { "role": "assistant", "content": "Check the operator controller's own pod logs, no...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A custom resource is stuck Terminating and won't actually delete." }, { "role": "assistant", "content": "A finalizer on the resource is waiting for something (usu...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Pods fail to create in a namespace with an error mentioning 'violates PodSecurity'." }, { "role": "assistant", "content": "Check the namespace's PodSecurity label...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "The whole cluster feels sluggish -- kubectl commands and pod scheduling are slow." }, { "role": "assistant", "content": "Check etcd health first, since it backs e...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl commands are timing out with 'connection refused' against the API server." }, { "role": "assistant", "content": "Check whether the apiserver's static pod ...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubeadm join fails with 'token has expired or is not valid'." }, { "role": "assistant", "content": "Bootstrap tokens expire after 24 hours by default -- generate ...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Upgraded the control plane and now some nodes show odd errors about unsupported API versions." }, { "role": "assistant", "content": "Kubernetes only supports kube...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Pods won't schedule onto a specific node pool even though it has capacity." }, { "role": "assistant", "content": "Check `kubectl describe node <node>` for a Taint...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Lower-priority pods keep getting evicted even though nothing looks wrong with them." }, { "role": "assistant", "content": "Check `kubectl get events --field-selec...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl drain hangs and never finishes evicting pods." }, { "role": "assistant", "content": "Check `kubectl get pdb -n <namespace>` -- drain respects PodDisruptio...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Pods stay Pending with 'did not match pod topology spread constraints' in the events." }, { "role": "assistant", "content": "Your `topologySpreadConstraints` with...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A Deployment can't scale past a certain number of replicas even though there's plenty of node capacity." }, { "role": "assistant", "content": "Check for `podAntiA...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "I have both a VerticalPodAutoscaler and HorizontalPodAutoscaler on the same deployment and scaling looks erratic." }, { "role": "assistant", "content": "VPA and H...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "HorizontalPodAutoscaler with a custom Prometheus metric shows <unknown> for the current value." }, { "role": "assistant", "content": "Query the custom metrics API...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Cluster autoscaler won't scale down a node even though it looks nearly empty." }, { "role": "assistant", "content": "Check for pods on that node with the `cluster...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A pod is stuck ContainerCreating with 'FailedAttachVolume' in events." }, { "role": "assistant", "content": "Check `kubectl describe pod <pod>` for the full event...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "PVCs across the whole cluster are stuck Pending and describe just says 'waiting for a volume to be created'." }, { "role": "assistant", "content": "Check the CSI ...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "I rotated a Secret's value but the pod using it as a volume mount still has the old one." }, { "role": "assistant", "content": "Same behavior as ConfigMaps: Secre...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "An app can't read its mounted config file even though the ConfigMap volume mount looks correct." }, { "role": "assistant", "content": "Check the volume's `default...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A pod shows ErrImageNeverPull even though the image is definitely present." }, { "role": "assistant", "content": "Check `imagePullPolicy` in the pod spec -- `Neve...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "My app gets intermittent 401s calling the Kubernetes API, but only after it's been running for a while." }, { "role": "assistant", "content": "Bound ServiceAccoun...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Pods fail to start with 'failed to create containerd task' in kubelet logs." }, { "role": "assistant", "content": "This is a container runtime-level failure, one ...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A node went NotReady with no workload changes, and kubelet logs mention a certificate error." }, { "role": "assistant", "content": "kubelet's client certificate r...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "After migrating nodes to a newer OS image, containers are getting OOMKilled at lower memory usage than before." }, { "role": "assistant", "content": "Newer OS ima...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A node hit DiskPressure and started evicting pods, but there's no obviously huge workload on it." }, { "role": "assistant", "content": "Unused container images an...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl top pod returns an error instead of numbers." }, { "role": "assistant", "content": "Check `kubectl get apiservices v1beta1.metrics.k8s.io` -- if it shows ...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl exec fails with 'error: unable to upgrade connection: container not found'." }, { "role": "assistant", "content": "This usually means the container name i...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl apply fails with 'the object has been modified; please apply your changes to the latest version and try again'." }, { "role": "assistant", "content": "Som...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "I applied a Kustomize overlay but the deployed resources don't have the patches I expected." }, { "role": "assistant", "content": "Render it locally first with `k...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "An ArgoCD Application keeps showing OutOfSync and re-syncing without ever settling." }, { "role": "assistant", "content": "Check `kubectl get application <name> -...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "helm install fails partway through and leaves resources behind." }, { "role": "assistant", "content": "Check `kubectl get pods -n <namespace> -l 'helm.sh/hook'` f...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "Ingress TLS isn't working and the cert-manager Certificate resource shows Ready: False." }, { "role": "assistant", "content": "`kubectl describe certificate <name...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "A Job's pod completed successfully but the Job itself shows Failed." }, { "role": "assistant", "content": "Check `kubectl describe job <name>` for `backoffLimit` ...

[ { "role": "system", "content": "You are a helpful Kubernetes and infrastructure expert." }, { "role": "user", "content": "kubectl delete namespace has been stuck in Terminating for a long time." }, { "role": "assistant", "content": "Almost always a leftover finalizer on some resource...