IOPS in Linux: Testing and optimising disk performance

What is IOPS?

Basic definition

IOPS (Input/Output Operations Per Second) – a metric that measures the number of input/output operations a storage device can perform in one second.

Think of IOPS as the speed at which your disk can:

• Read files
• Write new data
• Modify existing information

Why is IOPS important?

INFO

IOPS directly affects the speed of your applications and services

Critical use cases

• Databases (MySQL, PostgreSQL, MongoDB)
• High-load web servers
• File storage
• Caching systems
• Mail servers

Factors affecting IOPS

1. Drive type

• Used at F3 Cloud
• Maximum performance
• Up to 1,000,000 IOPS
• Minimal latency

2. Data block size

WARNING

Data block size significantly affects real-world performance

Block size	Impact on IOPS
4KB	Maximum IOPS
8KB	High IOPS
64KB	Medium IOPS
1MB	Low IOPS

3. Access patterns

• Random access
  • More demanding on IOPS
  • Typical of databases
  • Important for web applications
• Sequential access
  • Less demanding on IOPS
  • Characteristic of streaming
  • Important for backups

IOPS at F3 Cloud

TIP

On F3 Cloud servers, we do not set artificial limits on IOPS

Our advantages:

• Use of NVMe drives only
• No IOPS limits
• Optimised storage configuration
• Performance monitoring
• Ability to scale

How do you measure IOPS?

Linux

# Measuring random reads
fio --name=randread --ioengine=libaio --iodepth=16 --rw=randread --bs=4k --direct=1 --size=4G --numjobs=1 --runtime=240 --group_reporting

Windows

# Use CrystalDiskMark or ATTO Disk Benchmark

Optimisation recommendations

1. Choosing the right disk type
   • NVMe SSD for high-load systems
   • RAID arrays for improved reliability
2. File system configuration
   • Optimal block size
   • Correct mount options
3. Performance monitoring
   • Regular IOPS checks
   • Bottleneck analysis
   • Timely scaling

---

Our resources

Telegram channel: https://t.me/
F3 Cloud: https://f3cloud.com