[lkml]   [1999]   [Mar]   [31]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: softupdates and ext2
Trevor Johnson <> writes:

> > I made some benchs and it seems rather fast (about 5 to 10% slower than
> > the async mount, which is what Linux does).
> It does that by default, but (for ext2 at least) you can mount with the
> "sync" option to have at least the metadata written synchronously. It is
> much slower than the asynchronous way, but I would do it if the unclean
> shutdowns couldn't be prevented.

The reason the "sync" option is so slow is because it makes _all_ writes
synchronous. The default of sync metadata / async userdata used by most
Unixen is not noticeably slower than async except for large metadata
operations like removing large directory trees.

E.g. when testing with bonnie on a system using sync/async as default,
there is no visible difference from using "async". But when using "sync"
on a Linux system, the write performance dropped to ~80kbyte/s (on a
disk that writes ~8Mbytes/s using "async").

Mats Lofkvist

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
Please read the FAQ at

 \ /
  Last update: 2005-03-22 13:51    [W:0.017 / U:3.452 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site