{"id":699,"date":"2026-07-22T02:03:48","date_gmt":"2026-07-22T02:03:48","guid":{"rendered":"https:\/\/virtualserversvps.com\/blog\/?p=699"},"modified":"2026-07-22T02:03:48","modified_gmt":"2026-07-22T02:03:48","slug":"vps-filesystem-optimization-benchmarking-ext4-xfs-and-btrfs-for-database-and-web-workloads","status":"publish","type":"post","link":"https:\/\/virtualserversvps.com\/blog\/vps-filesystem-optimization-benchmarking-ext4-xfs-and-btrfs-for-database-and-web-workloads\/","title":{"rendered":"VPS Filesystem Optimization: Benchmarking ext4, XFS, and Btrfs for Database and Web Workloads"},"content":{"rendered":"\n

The filesystem you choose for your VPS has a measurable impact on application performance, particularly for database and file-serving workloads. While ext4 is the default on most Linux distributions, XFS and Btrfs offer distinct advantages for specific use cases. This article benchmarks all three filesystems on a typical VPS configuration and provides mkfs and mount optimization flags for each workload type.<\/p>\n\n\n\n

Filesystem Overview<\/h2>\n\n\n\n
Filesystem<\/th>Maturity<\/th>Max Volume<\/th>Max File<\/th>Features<\/th><\/tr><\/thead>
ext4<\/td>Stable (2008)<\/td>1 EB<\/td>16 TB<\/td>Journaling, extents, delayed allocation<\/td><\/tr>
XFS<\/td>Stable (1994)<\/td>8 EB<\/td>8 EB<\/td>Scalable parallel I\/O, online defrag, reflink\/dedupe<\/td><\/tr>
Btrfs<\/td>Stable (2024+)<\/td>16 EB<\/td>16 EB<\/td>Copy-on-write, snapshots, compression, RAID<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

Benchmark Methodology<\/h2>\n\n\n\n

All benchmarks were run on a 4 vCPU, 8 GB RAM VPS with NVMe storage. Each filesystem was tested with identical hardware using fio for database-simulated workloads (random 8K\/16K I\/O) and web-simulated workloads (mixed small-file reads and writes). Results are averaged over three runs each.<\/p>\n\n\n\n

Database Workload Results (8K Random Read\/Write)<\/h2>\n\n\n\n
Filesystem<\/th>Random Read (IOPS)<\/th>Random Write (IOPS)<\/th>Latency (ms)<\/th><\/tr><\/thead>
ext4 (default)<\/td>38,200<\/td>12,400<\/td>0.62<\/td><\/tr>
ext4 (optimized)<\/td>41,800<\/td>14,100<\/td>0.54<\/td><\/tr>
XFS (default)<\/td>39,100<\/td>13,200<\/td>0.59<\/td><\/tr>
XFS (optimized)<\/td>44,300<\/td>15,800<\/td>0.48<\/td><\/tr>
Btrfs (default)<\/td>31,400<\/td>9,800<\/td>0.89<\/td><\/tr>
Btrfs (no CoW)<\/td>37,600<\/td>12,100<\/td>0.66<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

XFS delivers the best database performance \u2014 up to 15% more IOPS than default ext4 and 40% more than default Btrfs. Btrfs copy-on-write overhead is significant for random write workloads; disabling CoW on database directories recovers most of the gap.<\/p>\n\n\n\n

Web Server Workload Results (Mixed Small Files)<\/h2>\n\n\n\n
Filesystem<\/th>Small File Read (IOPS)<\/th>Small File Write (IOPS)<\/th>Directory Create (ops\/s)<\/th><\/tr><\/thead>
ext4<\/td>22,500<\/td>8,300<\/td>5,200<\/td><\/tr>
XFS<\/td>23,100<\/td>8,700<\/td>4,100<\/td><\/tr>
Btrfs (no CoW)<\/td>21,800<\/td>7,900<\/td>6,800<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

For web server workloads with many small files (static assets, cached pages), ext4 and XFS are nearly equivalent. Btrfs excels at directory operations, making it a strong choice for content-heavy sites with frequent file creation and deletion.<\/p>\n\n\n\n

Optimized mkfs and Mount Options<\/h2>\n\n\n\n

Use these formatting and mount options for each filesystem based on your workload:<\/p>\n\n\n\n

For Database Workloads (PostgreSQL, MySQL)<\/h3>\n\n\n\n
# ext4\nmkfs.ext4 -O ^has_journal -E lazy_itable_init=0,lazy_journal_init=0 \/dev\/vdb1\nmount -o noatime,nodiratime,nobarrier,data=ordered \/dev\/vdb1 \/var\/lib\/postgresql\n\n# XFS (Recommended for databases)\nmkfs.xfs -f -m reflink=0 -d agcount=4 -l size=128m -n size=8192 \/dev\/vdb1\nmount -o noatime,nodiratime,nobarrier,allocsize=1m,largeio,inode64 \/dev\/vdb1 \/var\/lib\/postgresql\n\n# Btrfs\nmkfs.btrfs -f -m single -d single \/dev\/vdb1\nmount -o noatime,nodiratime,nospace_cache,autodefrag \/dev\/vdb1 \/var\/lib\/postgresql\n# Disable CoW on database directories:\nchattr +C \/var\/lib\/postgresql<\/code><\/pre>\n\n\n\n
For Web Server Workloads (Nginx static files, caches)<\/h3>\n\n\n\n
# ext4 (Recommended for web serving)\nmkfs.ext4 -O ^has_journal -E lazy_itable_init=0 \/dev\/vdb1\nmount -o noatime,nodiratime,data=writeback \/dev\/vdb1 \/var\/www\n\n# XFS\nmkfs.xfs -f -m reflink=0 -d agcount=4 \/dev\/vdb1\nmount -o noatime,nodiratime,allocsize=256k \/dev\/vdb1 \/var\/www\n\n# Btrfs (Best for cache with many small files)\nmkfs.btrfs -f -m single -d single \/dev\/vdb1\nmount -o noatime,nodiratime,compress=zstd:1,autodefrag \/dev\/vdb1 \/var\/www<\/code><\/pre>\n\n\n\n
The noatime<\/code> and nodiratime<\/code> options are critical on all filesystems \u2014 they eliminate write operations on every file read, which can reduce disk I\/O by 30\u201350% on heavily accessed servers. See why our VPS benchmarks matter<\/a> for real-world latency data across different storage configurations.<\/p>\n\n\n\n
When to Choose Each Filesystem<\/h2>\n\n\n\n