erofs-openbenchmark's People
Forkers
saumyapanda1erofs-openbenchmark's Issues
Slow EROFS creation speed with compression
Hello, I am a user of erofs and I appreciate your work on this project. I have been testing erofs on my Arch Linux system and I found that the erofs creation speed with compression is extremely slow compared to squashfs, which is another popular read-only file system. I would like to report this issue and request some improvement.
I used the following environment and tools to conduct the test:
- CPU: AMD 5800H 8c16t 3200 MHz
- RAM: 16 GB DDR4 3200 MHz
- SSD: TiPlus 5000 PCIe 3.0*4 2 TB
- OS: Arch Linux
- Kernel: 6.4.6
- erofs-utils: 1.6
- squashfs-tools: 4.6.1
I used a folder containing an Arch Linux container with Xfce desktop environment as the source for creating the file systems. The original size of the folder was about 3797448 KiB. I used 1 MiB as the block size for both squashfs and erofs.
I tested the creation speed and compression ratio of squashfs and erofs with different compression algorithms and levels. I also enabled deduplication for squashfs, which is supported by default, but not for erofs, which is still experimental. The results are shown in the following table:
File system | Compression algorithm | Compression level | Size (KiB) | Compression ratio | Time (s) | Speed (KiB/s) |
---|---|---|---|---|---|---|
squashfs | zstd | default (15) | 1436510.05 | 37.37% | 43.99 | 86328.69 |
squashfs | zstd | 6 | 1438565.88 | 37.88% | 16.77 | 226357.23 |
squashfs | zstd | 3 | 1493950.67 | 39.58% | 8.15 | 465885 |
squashfs | zstd | 1 | 1602424.36 | 42.20% | 6.96 | 545443 |
squashfs | lz4hc | default (9) | 1951430.30 | 51.39% | 7.37 | 514986 |
erofs | lz4 | default (1) | 3519608.00 | 92.68% | 10.24 | 370781 |
erofs | lz4hc | 9 | 1855644.00 | 48.47% | 388.97 | 9762 |
As you can see, erofs has a much slower creation speed which is more than 50 times slower than squashfs with the same compression algorithm and level (lz4hc @ level 9). This might be a significant drawback for using erofs, especially for large data sets.
I wonder if there is any way to improve the erofs creation speed with compression, or if there are any plans to do so in the future. I know that erofs is mainly designed for mobile phones or embedded devices which are of limited performance, but the creation may still be done on a higher performance device, and perhaps parallelization compression or other techniques are still necessary. I think erofs has a lot of potential as a read-only file system, but the slow creation speed may be an obstacle for its adoption.
I also want to acknowledge the advantages of erofs, such as its random access performance and layout design, which are very impressive and innovative. I hope that you can kindly consider my feedback and suggestions, and please let me know if you need any more information or assistance from me.
Thank you for your attention and response.
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. ๐๐๐
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google โค๏ธ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.