Visualize Performance Metrics
The System Resource Usage and performance charts in Pulse provide detailed insights into CPU, memory, and disk utilization for ClickHouse nodes.
Use these charts to:
- Identify CPU or memory bottlenecks.
- Detect I/O saturation and thread contention.
- Correlate hardware usage trends with query performance for proactive capacity planning.
Steps
- In the Pulse UI, go to ClickHouse > Dashboard.
- On the ClickHouse page, select the Performance tab.
- Set a time range and data refresh interval (for example, 10 s, 20 m, 2 h, 4 d, or 1 w) to view timely and relevant information.
Pulse then displays resource usage charts that provide a real-time view of system resource utilization and performance efficiency.
Charts
| Chart | Description |
|---|---|
| Total CPU User Time | Measures CPU time spent executing user-level (non-kernel) code by the ClickHouse server process, showing how much CPU capacity is used for query processing and computation. |
| Total CPU System Time | Measures CPU time spent in kernel mode on behalf of the ClickHouse process, including system calls and disk or network I/O. |
| Memory Usage | Shows total system RAM utilization, indicating how much physical memory is in use. High usage may signal memory pressure or caching overhead. |
| Disk Usage | Displays the percentage of disk space used on the storage volume where ClickHouse data resides. Useful for tracking growth and preventing capacity issues. |
| Total Threads | Indicates the total number of operating-system threads currently running on the host, including both ClickHouse and background processes. |
| Runnable Threads | Shows the number of threads ready to run (waiting in the CPU queue). High values may indicate CPU contention or saturation. |
| OS Context Switches | Measures how frequently the CPU switches between processes or threads. A high rate can signal contention for CPU resources. |
| OS Processes Blocked | Displays the number of processes currently blocked, typically waiting for I/O operations to complete. High values may indicate disk or network I/O bottlenecks. |
Key Features
Each chart includes the following options and actions:
- Show Anomaly Detection: Displays deviations from the normal pattern based on historical data.
- Tip: Refer to the chart description for more details.
- Refresh: Updates the chart to display the latest available data.
- Maximize: Expands the chart for a full-page view.
- Download: Exports the chart data in .xlsx or .csv format.
Was this page helpful?