How to use the command 'pveperf' (with examples)

The pveperf command is a performance benchmarking tool used in the Proxmox Virtual Environment (PVE). It assists system administrators in evaluating the performance of their Proxmox servers by gathering crucial metrics such as CPU and hard disk performance data. This tool is especially helpful in diagnosing performance issues and validating the capabilities of the server hardware. One of its primary applications is in plotting performance trends over time, which aids in proactive capacity planning and system optimization. By running pveperf, users obtain a comprehensive snapshot of their server’s performance parameters, which can then be analyzed for bottlenecks or inefficiencies.

Use case 1: Show CPU and hard disk performance data for the hard disk mounted at /

Code:

pveperf

Motivation:

The motivation for using this command lies in the need for system administrators to acquire instant insights into the system’s CPU and hard disk performance. This is crucial, especially after installing the Proxmox Virtual Environment or when there is a need to evaluate performance post hardware upgrades. By executing this command, administrators can identify current system capabilities and detect any potential performance bottlenecks that may hinder the performance of virtual machines running on the server. It helps in establishing baseline performance metrics, which are essential for comparative analysis over time as the workload on the server increases.

Explanation:

The pveperf command, when executed without any arguments, targets the default mount point, typically at /, to measure and display performance statistics of the CPU and the hard disk associated with that filesystem. By not specifying additional options, the user is requesting a straightforward analysis of system resources relevant to the base operating system and its immediate storage environment.

• pveperf: The command initiates the benchmark analysis. Running it without additional parameters means it evaluates the resources at the root filesystem, which is the common interest point for performance monitoring.

Example Output:

CPU BOGOMIPS:      63999.92
REGEX/SECOND:      1217102
HD SIZE:           95.32 GB (/dev/mapper/pve-root)
BUFFERED READS:    254.89 MB/sec
AVERAGE SEEK TIME: 7.89 ms
FSYNCS/SECOND:     2343.99
DNS EXT:           30.50 ms
DNS INT:           1.00 ms

In this example output, various metrics are revealed:

• CPU BOGOMIPS: A synthetic measurement that provides insight into the raw processing power of the CPU.
• REGEX/SECOND: Indicates how many regex operations can be done per second, offering insight into CPU string processing capability.
• HD SIZE: Displays the total size of the root mount disk.
• BUFFERED READS: Shows the reading speed from the buffer, reflecting the disk’s read performance.
• AVERAGE SEEK TIME: The time it takes for the disk to locate and begin reading a particular block of data.
• FSYNCS/SECOND: Measures how many sync operations the filesystem can perform per second, illustrating disk-write performance.
• DNS EXT/INT: Provides the latency in ms for external and internal DNS lookups, reflecting network responsiveness.

Conclusion:

The pveperf command is invaluable for system administrators who need to keep abreast of their Proxmox server’s performance. By using this tool, one gains essential visibility into how the physical resources are performing. This knowledge is foundational for optimizing the environment, supporting better troubleshooting methodologies, and ensuring that there are no hidden issues impacting system operations. Regular performance auditing with pveperf can lead to more predictable and reliable virtualized environments, thereby enhancing overall productivity and efficiency.