phoronix benchmarks ext4 vs. btrfs

Chris Murphy lists at
Fri Mar 9 08:15:54 UTC 2012

On Mar 9, 2012, at 12:30 AM, Matthias Runge wrote:
> if your file system places data inefficiently on disk/storage, you want
> to measure this, too. If you're comparing file system speed, I think,
> you should measure the whole thing and be sure to create comparable
> data.

The source files are on a jhfs+ volume. I'm disinterested in disk to disk copy. And I'm disinterested in comparing jhfs+ to other file system copies. I'm wondering how a system pushing data from memory to disk.

> You can't really control, how your file system cache is filled.
> Although I must say,  measuring a the whole effort a few times
> consecutively provides more reliable numbers (caution: but no
> disk speed measurement).

I can probably control it to some degree if there isn't much free memory left, by creating a large ramdisk, that will end up as file system cache.

Chris Murphy

More information about the devel mailing list