Visualize Node Details

The Nodes Details page provides interactive charts and summary details that help you visually analyze a specific node's resource usage and performance trends.

Steps

1. In the Pulse UI, select Nodes > Summary.

2. On the Nodes page, click the host name.

3. On the Node Details page:

   • Set time range – Select Today, Last 12 hours, Last 3 months, or choose a custom period. Then select Apply.
   • Service status: Click to view the current status (Up or Down) of all services running on the node.
   • Refresh status – Select Play (⏵) to refresh the page automatically every 10 seconds. Select Pause (⏸) to stop auto-refresh.

4. Review the summary panel and visualize the performance through charts.

Summary Panel

Review the summary panel, which displays key metrics across three categories:

Node Details

• Up Time – The duration the node has been running, which helps assess stability.
• CPUs – Number of processor cores available; shows processing capacity.
• Total Memory – Total physical memory on the node; baseline for usage tracking.

Resource Usage

• Memory Used – Percentage of memory in use; helps spot potential memory pressure.
• CPU Used – Percentage of CPU resources consumed; useful for detecting overload.
• Disk Used – Percentage of disk space utilized; aids in monitoring storage health.

Workload Distribution

• Applications – Total number of applications running on the node; shows workload volume.
• Containers – Number of YARN containers running; helps monitor distributed workloads.

Explore Charts

Explore detailed charts to analyze resource usage, workload distribution, and log trends and patterns across nodes.

Resource Usage

• CPU usage – Processing utilization on each node; helps track workload intensity.
• Memory usage – Trends for used, buffered, cached, and free memory; supports capacity planning.
• I/O usage – Operations per second, latency, read/write bytes, and request wait time; indicates storage performance.
• Network throughput – Sent and received bytes per second, with per-process drilldown; useful for detecting bottlenecks.
• Disk utilization – Usage across directories and mounts; helps identify space issues.

Workload Distribution

• Application resources – CPU and memory usage for the most intensive Hadoop processes; aids in performance tuning.
• YARN containers – Running containers and average launch duration; helps monitor workload distribution.

Logs & Diagnostics

• JVM garbage collection (GC): Displays garbage collection count and duration for core Hadoop services such as ResourceManager, DataNode, ZooKeeper, Kafka, and Hive Metastore to help debug memory problems.
• Error and fatal logs – Details by time, source, service, host, YARN App ID, and container ID; supports faster troubleshooting.

For detailed information about metrics and charts, see Node Details.