Dashboard Introduction

Grafana Dashboard is a visualization tool used to display monitoring data through charts, tables, and other visual forms. It extracts data from the data source Prometheus and organizes this data into a visual interface, helping users monitor and analyze system performance and status in real-time.

Node Monitoring

Data Source

Node monitoring is primarily based on Prometheus's node_exporter metrics.

Key Metrics

• CPU Usage: Displays real-time CPU usage for each node.
• Memory Usage: Used and remaining memory capacity for nodes.
• Disk I/O: Node disk read/write rate.
• Network Traffic: Inbound and outbound network traffic for each node.

K8S Monitoring

Data Source

K8S monitoring is based on Kubernetes's kubelet and cAdvisor metrics.

Key Metrics

• Pod Monitoring: Running status, restart count, resource allocation.

• Namespace Monitoring: Overall resource usage within namespaces.

• Workloads: Health status of Deployments, StatefulSets, DaemonSets.

MySQL Monitoring

Data Source

MySQL monitoring is based on metrics provided by mysqld_exporter.

Key Metrics

• QPS and TPS: Queries and transactions per second.
• Connections: Active connections and percentage of maximum connections.
• Slow Queries: Detect number of queries exceeding execution time threshold.
• Cache Hit Rate: Efficiency of InnoDB cache usage.

Redis Monitoring

Data Source

Redis monitoring is based on metrics provided by redis_exporter.

Key Metrics

• Hit Rate: Ratio of key request hits and misses.
• Memory Usage: Total memory used by Redis instance.
• Connections: Real-time number of client connections.
• Operation Rate: Number of commands executed per second.
• Master-Slave Status: Displays master-slave replication lag and health status.