The ack-node-problem-detector component is optimized and enhanced by Alibaba Cloud Container Service for Kubernetes (ACK) based on the open source Node Problem Detector (NPD) provided by the Kubernetes community. This component is used to monitor cluster nodes and integrate third-party node monitoring plug-ins. You can use ack-node-problem-detector to integrate custom monitoring plug-ins. This allows you to enhance the monitoring of nodes and detect more node anomalies.

For more information about the open source node-problem-detector, see node-problem-detector.

For more information about the installation, scenarios, and features of ack-node-problem-detector, see Event monitoring.

Release notes

Version Image address Release date Description
v0.6.3-28-160499f July 27, 2020
  • Adds the following information to out of memory (OOM) Killer events: the name of the relative pod, the namespace to which the pod belongs, and the user IDs (UIDs) of the killed processes.
  • Improves the efficiency of the check_fd plug-in.
  • Optimizes nodes events that reports that the process ID (PID) usage of cluster nodes exceeds the threshold.
  • Upgrades plug-ins that detect network connections.
  • Supports alert plug-ins that send alerts when the inode usage of the system disks of cluster nodes exceeds the threshold.