Back to articles
DevOps & Systems

Comprehensive Guide to Linux Performance Profiling on VPS Using Flamegraph

This guide provides a detailed walkthrough of Linux performance profiling on a VPS using the Flamegraph tool, an open-source visualization method for analyzing CPU usage. You will...

DPTCloud Engineering4 min read
Comprehensive Guide to Linux Performance Profiling on VPS Using Flamegraph

This guide provides a detailed walkthrough of Linux performance profiling on a VPS using the Flamegraph tool, an open-source visualization method for analyzing CPU usage. You will learn how to install the necessary tools, collect performance data, generate Flamegraph visualizations, and analyze performance bottlenecks on your VPS. The guide applies to common Linux systems such as Ubuntu, CentOS, and NVMe Cloud VPS platforms for Docker, Nginx, and Databases.

Introduction to Flamegraph and Linux Performance Profiling

Flamegraph is an open-source performance visualization tool developed by Brendan Gregg that enables you to visualize CPU usage by functions in a hierarchical, pyramid-shaped diagram. Each "layer" in the graph represents a function, and the width indicates execution time. Flamegraph helps quickly identify performance bottlenecks, especially in VPS environments running services like Nginx, Docker, or databases.

Linux performance profiling involves collecting data on system activity, including CPU, memory, I/O, and processing time. Tools such as perf, eBPF, and bpftrace gather raw data, while Flamegraph transforms this data into intuitive visualizations. Regular performance analysis ensures stable VPS operation, prevents overloads, and optimizes resource costs.

Prerequisites and Tool Installation

To begin, you need a Linux-based VPS (Ubuntu/Debian or CentOS/RHEL). Install the required tools as follows:

# Install necessary packages on Ubuntu/Debian
sudo apt update
sudo apt install -y git build-essential linux-tools-common linux-tools-generic linux-headers-generic

# Install perf (requires kernel debug symbols)
sudo apt install -y linux-image-$(uname -r)-dbgsym

# Install bpftrace (if needed)
sudo apt install -y bpftrace

# Clone and install Flamegraph from GitHub
git clone https://github.com/brendan Gregg/FlameGraph.git
cd FlameGraph

Note: Some VPS providers do not include kernel debug symbols. If you encounter debug symbol errors, check your VPS provider’s documentation or consider using NVMe Cloud VPS for Docker, Nginx & Databases to ensure full support.

Collecting Data and Generating Flamegraph

We will use perf to collect CPU data and convert it into a Flamegraph format. Here are sample commands:

# Collect CPU data for 10 seconds
perf record -g -F 99 -p $(pgrep nginx) -o perf.data -- sleep 10

# Convert data to Flamegraph format
cd FlameGraph
./stackcollapse-perf.pl perf.data > perf.folded

# Generate HTML flamegraph
./flamegraph.pl perf.folded > nginx-flamegraph.html

For non-service applications, replace $(pgrep nginx) with the process ID or use perf record -g -- to capture system-wide data.

For advanced performance tuning, refer to the guide on automated daily VPS backups to Cloudflare R2 using Borgmatic.

Analyzing and Optimizing Performance

In the Flamegraph, functions with wider bars consume more CPU time. When you open nginx-flamegraph.html in a browser, focus on:

  • Functions with large widths: high CPU consumption.
  • Long function chains: possible loops or repeated calls.
  • System functions like malloc, memcpy: potential memory or I/O issues.

After identifying bottlenecks, optimize by:

  • Reducing unnecessary requests.
  • Optimizing code or using more efficient libraries.
  • Adjusting service configurations (e.g., Nginx worker count).
  • Scaling up VPS resources if needed.

Testing and Verification

To verify changes, repeat the data collection process and compare pre- and post-optimization Flamegraphs. Tools like htop, top, or vmstat can also help confirm performance improvements.

Not all VPS environments fully support perf. If you encounter issues, verify kernel compatibility, install debug symbols, or consider using High-Performance Dedicated Servers for Production Clusters.

Troubleshooting

  • No data collected: Ensure perf is installed and running with sufficient permissions. Run perf list to check supported events.
  • \"No symbols found\" error: Install kernel debug symbols or use the correct kernel version.
  • Graph not displaying: Verify that flamegraph.pl executes correctly and has no syntax errors.

Refer to the official documentation at https://github.com/brendan Gregg/FlameGraph.

DPTCloud Tier III Infrastructure

Recommended NVMe Cloud VPS Plans to Deploy This Setup

Enterprise RAID 10 NVMe SSDs, CMC & FPT District 9 uplinks with Anycast Anti-DDoS and 99.99% SLA.

View 22 VPS Plans
ECO · vn-02

VPS-2E

94.600 ₫ / mo

10% VAT included

  • CPU: 2 vCPU
  • RAM: 2 GB
  • SSD: 30 GB NVMe
ECO · vn-02Recommended

VPS-3E

170.500 ₫ / mo

10% VAT included

  • CPU: 2 vCPU
  • RAM: 4 GB
  • SSD: 50 GB NVMe
BASIC · vn-01Recommended

VPS-05

457.600 ₫ / mo

10% VAT included

  • CPU: 4 vCPU
  • RAM: 8 GB
  • SSD: 50 GB NVMe

Related Articles

View all
ZaloZalo SupportDiscord0384 753 478