This article explains why a Huawei OBS CSI plugin loses its s3fs process after a restart, causing "Transport endpoint is not connected" errors, and provides a step‑by‑step solution using client‑go to rebuild the mount and trigger kubelet remount via a liveness probe.
The article investigates a cloud‑vendor Kubernetes isolation feature that inserts iptables DROP rules into a pod’s network namespace, demonstrating how it fully blocks traffic, triggers liveness‑probe restarts, and impacts services depending on replica count and probe configuration, while preserving state only without probes.
This article explains why Kubernetes uses Startup, Liveness, and Readiness probes, describes common Pod states and restart policies, compares the probes, details their configuration fields, and provides practical YAML examples for each probe type to ensure reliable container health monitoring.
This article explains why Kubernetes needs Startup, Liveness, and Readiness probes, details pod lifecycle states, restart policies, probe configurations, and provides practical YAML examples for each probe type to ensure reliable service availability during updates and failures.
This guide explains how Kubernetes uses livenessProbe and readinessProbe to monitor container health, prevent traffic loss, and trigger restarts, covering probe types, implementation methods, configuration parameters, and practical examples with exec, HTTP GET, and TCP socket actions.
