monitoring 归档 - 夜莺博客

node-problem-detector: Node Health the Cluster Can See

Deploy node-problem-detector to surface kernel, kubelet and hardware problems as Kubernetes events, conditions and Prometheus metrics.

admin 2026-10-14

0 0 0

Network Time Security with Chrony (RFC 8915)

Replace unauthenticated NTP with NTS: configure chrony for Network Time Security, verify the KE handshake and cookies, and keep time reliable under at

admin 2026-10-13

0 0 0

Prometheus Alertmanager Routing and Inhibition Rules

Design Alertmanager routing trees and inhibition rules: group_by and timing choices, child routes with continue, and matchers that stop alert storms.

admin 2026-10-13

0 0 0

syslog-ng Relay Architecture for Network Logs

Build a syslog-ng relay tier: source, filter, destination and log path syntax, disk buffering for outages, and TLS between relays and the central serv

admin 2026-10-13

0 0 0

Prometheus snmp_exporter Alerting for Network Devices

Build network device monitoring with snmp_exporter's generator and Prometheus rules: generator.yml modules, walk parameters, and alert expressions tha

admin 2026-10-13

0 0 0

SNMP Traps vs Polling: Designing Network Alerting

SNMP traps vs polling for network alerting: which events need push, poll intervals that make sense, alarm deduplication and avoiding alert storms.

admin 2026-10-10

0 0 0

rsyslog for Network Device Logs: Reliable Collection

Collect switch and router logs with rsyslog: TCP and TLS listeners, per-device files, rate limits, parsing, rotation and retention that survives audit

admin 2026-10-10

0 0 0