AWS
339 incident problems in AWS environments.
먼저 읽을 가이드
추천 문제
All problems (339)
K8S-1268An ALB target group stays unhealthyA newly exposed service is reachable in a browser but the load balancer continues to mark every target unhealthy.KubernetesIntermediate16 minProK8S-1258An HPA stays at unknown metricsA team enables autoscaling and later discovers CPU or memory metrics never populate on an otherwise healthy deployment.KubernetesIntermediate16 minProK8S-1267Cluster autoscaler sees unschedulable pods but never scalesA cluster with pending workloads never scales up even though the autoscaler deployment itself appears healthy.KubernetesIntermediate16 minProK8S-1272CoreDNS runs but pods resolve external names incorrectlyAn EKS cluster starts returning odd external DNS results after a VPC-level network change, while CoreDNS remains Ready.KubernetesIntermediate16 minProK8S-1274Metrics Server is installed but HPAs stay blindAn HPA remains at unknown metrics even after Metrics Server was installed and the pods appear healthy.KubernetesIntermediate16 minProK8S-393A multi-cluster failover record updates correctly while clients pinned to one HTTP/2 connection never rediscover the new backend set during a staged decommissionName resolution converges and long-lived streams keep the old world alive. The service still works through the primary path, but one dependency only fails when the old component is finally drained away.KubernetesAdvanced17 minProCICD-294A package provenance check validates one tarball while the deploy step consumes a repacked archive from a mirror during a failover rehearsalSupply-chain checks pass on the source artifact, but the runtime path uses a derived archive that was never verified. Normal traffic masked the issue until the standby or alternate path became active under rehearsal conditions.CI/CDIntermediate17 minProK8S-357A PersistentVolume reattaches (Rollout Stuck)A PersistentVolume reattaches (Rollout Stuck) focuses on kubernetes-storage-and-state and asks the reader to isolate Rollout Stuck in AWS. 실무에서는 rollout-stuck 증상만 보고 Pod 하나에 매달리지 말고 이벤트, 이전 로그, Service/Endpoint, 최근 배포 변경을 한 번에 묶어 보는 편이 오진을 줄입니다.KubernetesAdvanced17 minProCICD-306A preview deployment signs in with a lower-privilege identity and the later promote step assumes those same credentials are still sufficient during a failover rehearsalThe early environment works, but the promotion path needs a broader scope that the pipeline never refreshes. Normal traffic masked the issue until the standby or alternate path became active under rehearsal conditions.CI/CDIntermediate17 minProK8S-1260A Service exposes the app but external traffic still failsA Service exposes the app but external traffic still fails focuses on firewall-policy-basics and asks the reader to isolate the key signal in AWS. A healthy Service and healthy pods do not prove the cloud load balancer can traverse the node-s...KubernetesIntermediate17 minProCICD-192An artifact retention job keeps the image and removes the detached attestation during a failover rehearsalThe primary artifact remains available while admission or provenance checks lose the supporting proof they depend on. Normal traffic masked the issue until the standby or alternate path became active under rehearsal conditions.CI/CDIntermediate17 minProK8S-1257An EKS ingress stays pendingAn ingress or LoadBalancer Service never provisions even though the controller is installed and has permissions.KubernetesIntermediate17 minProK8S-137An external load balancer health check hits the NodePort before readiness gating is effective, and nodes enter rotation with no viable backendsInfrastructure sees open ports, but the application is not actually ready to serve traffic yet.KubernetesAdvanced17 minProCICD-276Preview environment teardown removes DNS before certificate validation jobs finish probing the old hostname during a failover rehearsalThe environment lifecycle cleanup is correct for cost control and too early for the remaining verification path. Normal traffic masked the issue until the standby or alternate path became active under rehearsal conditions.CI/CDIntermediate17 minProCICD-487A preview environment uses feature flags from production defaults and hides...A preview environment uses feature flags from production defaults and hides... focuses on ci-cd-workflow-debugging and asks the reader to isolate Permission Denied in AWS. 실무에서는 ci-cd-workflow-debugging 문제를 볼 때 실패 단계만 보지 말고 최근 변경, 이미지 태그, 시크릿 주입, 롤백 가능 여부를 먼저 함께 확인하는 편이 빠릅니다.CI/CDIntermediate18 minProCICD-222A promotion rule trusts semantic version metadata while mirror sync rewrites the build metadata during a failover rehearsalThe release artifact keeps its version label, yet the mirror path mutates the metadata field one downstream gate still uses. Normal traffic masked the issue until the standby or alternate path became active under rehearsal conditions.CI/CDAdvanced18 minProK8S-123A restored PersistentVolume keeps node affinity for the old zone and the replacement pod never mounts in the new region segmentRestore succeeded in storage terms, but scheduling and attach logic still reflect the source topology.KubernetesAdvanced18 minProK8S-264A source-range rule trusts node addresses while externalTrafficPolicy Local leaves one zone with no eligible ingress path during a failover rehearsalThe load balancer policy is correct and locality changes which nodes can actually receive traffic. Normal traffic masked the issue until the standby or alternate path became active under rehearsal conditions.KubernetesAdvanced18 minProK8S-1235An init job fixes permissions but the main container still failsA pod initializes successfully on some restarts and fails on others depending on the relative timing of the CSI mount and init logic.KubernetesIntermediate18 minProCICD-547A preview environment uses feature flags from production defaults and hides...A preview environment uses feature flags from production defaults and hides... focuses on ci-cd-workflow-debugging and asks the reader to isolate Permission Denied in AWS. 실무에서는 ci-cd-workflow-debugging 문제를 볼 때 실패 단계만 보지 말고 최근 변경, 이미지 태그, 시크릿 주입, 롤백 가능 여부를 먼저 함께 확인하는 편이 빠릅니다.CI/CDIntermediate19 minPro