Observability Pipeline
181 incident problems about Observability Pipeline. Start with the reviewed ones.
먼저 읽을 가이드
추천 문제
All problems (181)
K8S-1409A host-networked observability daemon reaches its collectors and later loses...A host-networked observability daemon reaches its collectors and later... focuses on Observability Pipeline and asks the reader to isolate the key signal in cilium. Host-network pods often rely on policy exceptions that do not automaticall...KubernetesAdvanced14 minProK8S-1310Grafana Agent remote_write looks connected but drops a burst of cluster metricsMetrics ingestion resumes after pressure recovery, but dashboards never fill the missing historical gap from the outage window.KubernetesIntermediate14 minProK8S-1306Promtail tails files successfully until log rotation acceleratesA noisy cluster starts losing short-lived container logs even though Promtail itself reports no crash or auth error.KubernetesIntermediate14 minProLINUX-1523A logrotate run finishes and disk stays fullDisk usage never drops after log rotation on a host with a containerized writer.LinuxIntermediate9 minProLINUX-1533A logrotate run finishes and disk stays fullDisk usage never drops after log rotation on a host with a containerized writer.LinuxIntermediate9 minProLINUX-1503A logrotate run succeeds and disk usage stays highDisk usage never falls after rotation on a host with containerized log writers.LinuxIntermediate9 minProLINUX-1513A rotated log file disappears and disk stays fullDisk usage never drops after rotation on a host running a containerized app with bind-mounted logs.LinuxIntermediate9 minProK8S-1426A Grafana Agent deployment keeps one namespace darkA Grafana Agent deployment keeps one namespace dark focuses on Observability Pipeline and asks the reader to isolate the key signal in grafana. Operator watch filters can exclude newly labeled namespaces for longer than teams expe...KubernetesIntermediate10 minProK8S-1436A Grafana Agent operator misses one namespaceA Grafana Agent operator misses one namespace focuses on Observability Pipeline and asks the reader to isolate the key signal in grafana. Operator onboarding gaps can come from late label injection missing the first informer snaps...KubernetesIntermediate10 minProLINUX-1493A logrotate run completes and disk usage stays highDisk usage never falls after log rotation on a host where an app writes through a bind-mounted log path from a container.LinuxIntermediate10 minProLINUX-1419An rsyslog failover action stamps overflow files with the wrong hostnameAn rsyslog failover action stamps overflow files with the wrong hostname focuses on Observability Pipeline and asks the reader to isolate the key signal in redhat. Fallback logging branches often preserve assumptions from older init s...LinuxIntermediate10 minProLINUX-1439An rsyslog overflow file shows a stale hostnameAn rsyslog overflow file shows a stale hostname focuses on Observability Pipeline and asks the reader to isolate the key signal in redhat. Fallback logging branches often preserve legacy environment assumptions after the main path...LinuxIntermediate10 minProLINUX-1429An rsyslog overflow file shows the wrong hostnameAn rsyslog overflow file shows the wrong hostname focuses on Observability Pipeline and asks the reader to isolate the key signal in redhat. Fallback logging branches often preserve old environment assumptions long after the primary...LinuxIntermediate10 minProLINUX-1409rsyslog writes normal logs in UTC and fallback overflow logs in local timersyslog writes normal logs in UTC and fallback overflow logs in local time focuses on Observability Pipeline and asks the reader to isolate the key signal in redhat. Time skew symptoms in logs can be formatting-path bugs rather than actu...LinuxIntermediate10 minProK8S-1416A Grafana Agent DaemonSet stays healthy and one node never forwards logsA Grafana Agent DaemonSet stays healthy and one node never forwards logs focuses on Observability Pipeline and asks the reader to isolate the key signal in grafana. DaemonSet logging issues often come from node-image-specific log pat...KubernetesIntermediate11 minProK8S-1406A Grafana Agent operator rollout leaves one namespace unmonitoredA Grafana Agent operator rollout leaves one namespace unmonitored focuses on Observability Pipeline and asks the reader to isolate the key signal in grafana. Operator watch caches can miss freshly added selector labels for longer than teams...KubernetesIntermediate11 minProK8S-1441A PodMonitor stops matching one tenant after a post-render step strips the release label only in that namespaceTargets are healthy, but one rendered namespace no longer satisfies the PodMonitor selector.KubernetesIntermediate11 minProK8S-1421A Prometheus Operator rollout keeps one workload uns scrapedA Prometheus Operator rollout keeps one workload uns scraped focuses on Observability Pipeline and asks the reader to isolate the key signal in Prometheus. Prometheus scraping gaps often come from per-environment render differe...KubernetesIntermediate11 minProK8S-1431A Prometheus Operator upgrade leaves one workload darkOne tenant loses monitoring after a Helm post-renderer rollout while every other namespace continues to scrape correctly.KubernetesIntermediate11 minProK8S-1411A Prometheus scrape target disappears (podmonitor-selector-relied-on-legacy-mirrored-namespace-label)A Prometheus scrape target disappears (podmonitor-selector-relied-on-legacy... focuses on Observability Pipeline and asks the reader to isolate the key signal in Prometheus. Monitoring selectors can depend on label mirroring behav...KubernetesIntermediate11 minPro