4 ms·
> For I/O-heavy workloads with a lot of metadata changes - unpacking a tarball was given as an example - Dave said that ext4 could be 20-50 times faster than XF
by kijin 15y ago
> For I/O-heavy workloads with a lot of metadata changes - unpacking a tarball was given as an example - Dave said that ext4 could be 20-50 times faster than XFS.
A while ago, I read somewhere that XFS is good for a small number of large files, whereas ReiserFS is good for a large number of small files. I don't know whether that's true anymore if it ever was; but perhaps extracting a tarball with thousands of small files in it is not the best way to bring out XFS's strength?
By the way, when are we expecting the "experimental" label to be taken off of btrfs?
- noahdesu 15y ago> By the way, when are we expecting the "experimental" label to be taken off of btrfs? I believe distributions are waiting on an fsck for btrfs.
- Tuna-Fish 15y ago> A while ago, I read somewhere that XFS is good for a small number of large files, whereas ReiserFS is good for a large number of small files. I don't know whether that's true anymore if it ever was It used to be true. XFS has for a long time been the very fastest real filesystem available when it comes to moving data, and one of the slowest ones in real-world use when it comes to changing metadata. Untarring a kernel tarball a second time (on top of the first) is pretty much the canonical benchmark for metadata operations, and the one where XFS used to be at it's absolute worst. XFS is in the news right now exactly because they have fixed that deficiency. It's still faster than ext*, zfs and btrfs in pretty much everything else, but now it's comparable to them when modifying metadata. So this is exactly the right benchmark to use, because it's the one where there has been recent change. Video of Dave Chinner presenting this at linux.conf.au: https://www.youtube.com/watch?v=FegjLbCnoBw https://www.youtube.com/watch?v=FegjLbCnoBw