EXT3 vs EXT4 vs XFS vs BTRFS filesystem comparison on Fedora 18
Page 3 of 10
FS_MARK
Benchmark profile: - 1 thread / 8192 files 4K each - 32 threads / 8192 files 4K each
A new entry of my bench suite is fs_mark. Let see how the different filesystems perform when creating many small files (8 KB in size), with 1 and 32 threads.
EXT4 is the leader here, but note XFS's good show: with its complex metadata and journal system, this result is quite significant. This good score can be attributed to the “delayed logging” feature. Moreover, it consume very low CPU time (not that the other are too much CPU-heavy, anyway). BTRFS multi-thread score is quite low relative to the others, however it remain very high in absolute terms.
Let also see synchronized (fsync) write speed:
The number are now much lower, but the standing remain more or less the same.
Comments
It would be very interesting for future tests, how ZFS on Linux performs. Especially, after it has become "productive" some weeks ago.
this is surely a good idea ;)
I will investigate this possibility for the next review.
Regards.
You don't need to set the entire filesystem to be non COW, only the directory and files that need it (chattr -C flag).
Hi Ivan,
you are right. Anyway, I benchmarked BTRFS even with disabled CoW and found that, for virtual machines at least, it performs noticeably worse than a traditional filesystem as EXT4.
You can read more here:
http://www.ilsistemista.net/index.php/linux-a-unix/36-btrfs-mount-options-and-virtual-machines-an-in-depth-look.html
The only catch is that both tests are somewhat old now, being performed on Fedora 17 and 18. I should really see if with newer kernels BTRFS performances are better now.
But I have so little time ;)
RSS feed for comments to this post