On Mon, Dec 1, 2014 at 4:54 PM, Jan Kara <jack@xxxxxxx> wrote:
> Hello,
>
> On Mon 01-12-14 12:32:05, Zhi Yong Wu wrote:
>> I'd like to attend LSF/MM summit, been very interested in the
>> discussion related to VFS, ext4, xfs, btrfs, and memory management;
>> and would like to discuss some FS perf stuff with others.
> Can you be more specific what performance discussions are you interested
> in? Thanks.
e.g.
1. resize2fs/resize4fs is one of them.
When ext4 fs is resized by a big size, you can see that its performance
is much worse than windows ntfsresize.
Below is some perf data:
1.) single fsresize operation
orig_fssize -> new_fsresize ext4 ntfs
(mm:ss)
71680M -> 174080M 0:43.96 0:14.43
174080M -> 276480M 0:47.52 0:13.66
276480M -> 378880M 0:51.13 0:13.93
378880M -> 481280M 0:56.96 0:13.40
481280M -> 1505280M 5:57.19 0:32.29
2.) 4 fsresize operation in parallel
orig_fssize -> new_fsresize ext4 ntfs
(mm:ss)
174080M -> 276480M 1:05.83 0:38.87
276480M -> 378880M 1:08.39 0:34.69
378880M -> 481280M 1:12.55 0:35.26
481280M -> 1505280M 7:35.15 0:40.12
2. another perf issue is being resolved by Ted's lazytime patchset. :)
>
> Honza
> --
> Jan Kara <jack@xxxxxxx>
> SUSE Labs, CR
--
Regards,
Zhi Yong Wu
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
