I am running Ubuntu 22.04.2 on a laptop with the following specifications:
SSD: HP S700 SSD
CPU: Intel Core i7 7700HQ
RAM: 16GB
Kernel: 5.19.0-32-generic
All system packages are updated to the latest versions as of 25/02/2023. Ubuntu is installed on an Ext4 partition, while my data is stored on an NTFS partition since it was created when using Windows 10. The SSD is recognized as /dev/sdb
Details of the SSD and partitions:
Output of hdparm -I /dev/sdb:
/dev/sdb:
ATA device, with non-removable media
Model Number: HP SSD S700 1TB
Serial Number: HASA42470101207
Firmware Revision: V0823A0
Media Serial Num:
Media Manufacturer:
Transport: Serial, ATA8-AST, SATA 1.0a, SATA II Extensions, SATA Rev 2.5, SATA Rev 2.6, SATA Rev 3.0
Standards:
Used: unknown (minor revision code 0x011b)
Supported: 10 9 8 7 6 5
Likely used: 10
Configuration:
Logical max current
cylinders 16383 16383
heads 16 16
sectors/track 63 63
--
CHS current addressable sectors: 16514064
LBA user addressable sectors: 268435455
LBA48 user addressable sectors: 1953525168
Logical Sector size: 512 bytes
Physical Sector size: 512 bytes
Logical Sector-0 offset: 0 bytes
device size with M = 1024*1024: 953869 MBytes
device size with M = 1000*1000: 1000204 MBytes (1000 GB)
cache/buffer size = unknown
Form Factor: 2.5 inch
Nominal Media Rotation Rate: Solid State Device
Capabilities:
LBA, IORDY(can be disabled)
Queue depth: 32
Standby timer values: spec'd by Standard, no device specific minimum
R/W multiple sector transfer: Max = 1 Current = 1
Advanced power management level: 254
DMA: mdma0 mdma1 mdma2 udma0 udma1 udma2 udma3 udma4 udma5 *udma6
Cycle time: min=120ns recommended=120ns
PIO: pio0 pio1 pio2 pio3 pio4
Cycle time: no flow control=120ns IORDY flow control=120ns
Commands/features:
Enabled Supported:
* SMART feature set
Security Mode feature set
* Power Management feature set
* Write cache
* Look-ahead
* Host Protected Area feature set
* WRITE_BUFFER command
* READ_BUFFER command
* NOP cmd
* DOWNLOAD_MICROCODE
* Advanced Power Management feature set
* 48-bit Address feature set
* Mandatory FLUSH_CACHE
* FLUSH_CACHE_EXT
* General Purpose Logging feature set
* WRITE_{DMA|MULTIPLE}_FUA_EXT
* {READ,WRITE}_DMA_EXT_GPL commands
* Segmented DOWNLOAD_MICROCODE
* Gen1 signaling speed (1.5Gb/s)
* Gen2 signaling speed (3.0Gb/s)
* Gen3 signaling speed (6.0Gb/s)
* Native Command Queueing (NCQ)
* Phy event counters
* READ_LOG_DMA_EXT equivalent to READ_LOG_EXT
* DMA Setup Auto-Activate optimization
* Software settings preservation
* SMART Command Transport (SCT) feature set
* SANITIZE feature set
* BLOCK_ERASE_EXT command
* DOWNLOAD MICROCODE DMA command
* WRITE BUFFER DMA command
* READ BUFFER DMA command
* Data Set Management TRIM supported (limit 8 blocks)
Security:
Master password revision code = 65534
supported
not enabled
not locked
frozen
not expired: security count
supported: enhanced erase
6min for SECURITY ERASE UNIT. 6min for ENHANCED SECURITY ERASE UNIT.
Checksum: correct
Output of df -aTh:
Filesystem Type Size Used Avail Use% Mounted on
/dev/sdb2 ext4 162G 64G 90G 42% /
/dev/sdb1 fuseblk 760G 383G 378G 51% /mnt/DATA Laptop
Output of findmnt --fstab:
TARGET SOURCE FSTYPE OPTIONS
/ UUID=79debb50-d530-d801-70da-b150d530d801 ext4 errors=remount-ro
/mnt/DATA Laptop UUID=01D94304BCDF47E0 ntfs uid=1000,gid=1000,rw,user,exec,umask=000,x-gvfs-show
Performance difference between NTFS and Ext4 partitions:
When I ran a benchmark using KDiskMark on both partitions, I noticed a significant difference in performance between the NTFS and Ext4 partitions, especially for random read writes.
Output for /dev/sdb2:
KDiskMark (3.1.2): https://github.com/JonMagon/KDiskMark
Flexible I/O Tester (fio-3.28): https://github.com/axboe/fio
--------------------------------------------------------------------------------
* MB/s = 1,000,000 bytes/s [SATA/600 = 600,000,000 bytes/s]
* KB = 1000 bytes, KiB = 1024 bytes
[Read]
Sequential 1 MiB (Q= 8, T= 1): 517.054 MB/s [ 504.9 IOPS] < 15813.82 us>
Sequential 1 MiB (Q= 1, T= 1): 500.349 MB/s [ 488.6 IOPS] < 2044.34 us>
Random 4 KiB (Q= 32, T= 1): 168.011 MB/s [ 42002.8 IOPS] < 760.80 us>
Random 4 KiB (Q= 1, T= 1): 28.314 MB/s [ 7078.7 IOPS] < 139.87 us>
[Write]
Sequential 1 MiB (Q= 8, T= 1): 356.269 MB/s [ 347.9 IOPS] < 24887.09 us>
Sequential 1 MiB (Q= 1, T= 1): 334.044 MB/s [ 326.2 IOPS] < 2978.64 us>
Random 4 KiB (Q= 32, T= 1): 184.879 MB/s [ 46219.8 IOPS] < 665.75 us>
Random 4 KiB (Q= 1, T= 1): 84.579 MB/s [ 21144.9 IOPS] < 46.37 us>
Profile: Default
Test: 1 GiB (x5) [Measure: 5 sec / Interval: 5 sec]
Date: 2023-02-25 14:18:20
OS: ubuntu 22.04 [linux 5.19.0-32-generic]
Output for /dev/sdb1:
KDiskMark (3.1.2): https://github.com/JonMagon/KDiskMark
Flexible I/O Tester (fio-3.28): https://github.com/axboe/fio
--------------------------------------------------------------------------------
* MB/s = 1,000,000 bytes/s [SATA/600 = 600,000,000 bytes/s]
* KB = 1000 bytes, KiB = 1024 bytes
[Read]
Sequential 1 MiB (Q= 8, T= 1): 484.235 MB/s [ 472.9 IOPS] < 16879.14 us>
Sequential 1 MiB (Q= 1, T= 1): 494.241 MB/s [ 482.7 IOPS] < 2068.27 us>
Random 4 KiB (Q= 32, T= 1): 24.336 MB/s [ 6084.1 IOPS] < 5255.13 us>
Random 4 KiB (Q= 1, T= 1): 24.453 MB/s [ 6113.4 IOPS] < 162.18 us>
[Write]
Sequential 1 MiB (Q= 8, T= 1): 313.461 MB/s [ 306.1 IOPS] < 8380.51 us>
Sequential 1 MiB (Q= 1, T= 1): 313.816 MB/s [ 306.5 IOPS] < 879.35 us>
Random 4 KiB (Q= 32, T= 1): 78.391 MB/s [ 19597.8 IOPS] < 1237.62 us>
Random 4 KiB (Q= 1, T= 1): 77.982 MB/s [ 19495.5 IOPS] < 37.17 us>
Profile: Default
Test: 1 GiB (x5) [Measure: 5 sec / Interval: 5 sec]
Date: 2023-02-25 14:14:16
OS: ubuntu 22.04 [linux 5.19.0-32-generic]
Given that the NTFS partition is being recognized as a fuseblk system by df, I assume that the partition is being mounted using ntfs-3g instead of ntfs3, which is the new kernel driver. Will changing my fstab FSTYPE to ntfs3 as described here improve performance? Are there any other steps I can take to improve NTFS performance on Linux? If not, are there any ways to non-destructively convert an NTFS partition to Ext4?
Thank you in advance for your help and have a nice day.
Update:
Setting the FSTYPE to ntfs3 in /etc/fstab did indeed improve the performance a lot.
Output of KDiskMark for /dev/sdb1:
KDiskMark (3.1.2): https://github.com/JonMagon/KDiskMark
Flexible I/O Tester (fio-3.28): https://github.com/axboe/fio
--------------------------------------------------------------------------------
* MB/s = 1,000,000 bytes/s [SATA/600 = 600,000,000 bytes/s]
* KB = 1000 bytes, KiB = 1024 bytes
[Read]
Sequential 1 MiB (Q= 8, T= 1): 525.712 MB/s [ 513.4 IOPS] < 15556.55 us>
Sequential 1 MiB (Q= 1, T= 1): 428.302 MB/s [ 418.3 IOPS] < 2387.84 us>
Random 4 KiB (Q= 32, T= 1): 237.324 MB/s [ 59331.2 IOPS] < 538.66 us>
Random 4 KiB (Q= 1, T= 1): 43.243 MB/s [ 10810.8 IOPS] < 92.45 us>
[Write]
Sequential 1 MiB (Q= 8, T= 1): 495.466 MB/s [ 483.9 IOPS] < 16344.70 us>
Sequential 1 MiB (Q= 1, T= 1): 372.929 MB/s [ 364.2 IOPS] < 2591.59 us>
Random 4 KiB (Q= 32, T= 1): 299.469 MB/s [ 74867.4 IOPS] < 425.59 us>
Random 4 KiB (Q= 1, T= 1): 135.961 MB/s [ 33990.3 IOPS] < 27.96 us>
Profile: Default
Test: 1 GiB (x5) [Measure: 5 sec / Interval: 5 sec]
Date: 2023-02-25 15:45:46
OS: ubuntu 22.04 [linux 5.19.0-32-generic]
However, if I change the mount option to ntfs3, I am unable to play any steam games on the partition which rely on proton. I can only run those games when I use the ntfs mount option which uses ntfs-3g.
Update 2:
I found a workaround to use ntfs3 with proton, so I will be keeping that flag in /etc/fstab. Any other suggestions for NTFS performance improvements or filesystem conversions from NTFS which are both non-destructive and compatible with both Windows 10 and Linux are welcome.
ntfs3driver in the kernel, which seems performant enough for my needs. Are there any guides on mounting options or kernel configurations to further optimize the driver?I found a workaround to use ntfs3 with proton