Stephen Tweedie wrote:
>
> Hi,
>
> On Tue, Jun 25, 2002 at 10:03:47AM -0700, Andrew Morton
> <[email protected]> wrote:
>
> > If it's because of the disk-spins-up-too-much problem then
> > that can be addressed by allowing the commit interval to be
> > set to larger values.
>
> > +int jbd_commit_interval = 5; /* /proc/sys/fs/jbd_commit_interval */
>
> I suspect you want this to be per-mount, not system-wide (although
> filesystems could easily just inherit the system default dynamically
> if there's no per-fs override.) I could easily imagine a user wanting
> a different interval for a scratch disk, for example.
>
Yes, that would be better. We do want to be able to change it
on the fly. So how about:
mount /dev/what /mnt/where -o commit_interval=5
and
mount /mnt/where -o remount,commit_interval=3000
?
-
Hi,
On Wed, Jun 26, 2002 at 11:39:31AM -0700, Andrew Morton wrote:
> Yes, that would be better. We do want to be able to change it
> on the fly. So how about:
>
> mount /dev/what /mnt/where -o commit_interval=5
> and
> mount /mnt/where -o remount,commit_interval=3000
I just implemented that on the way back from OLS, will check it in
shortly.
--Stephen
Hi!
> > Yes, that would be better. We do want to be able to change it
> > on the fly. So how about:
> >
> > mount /dev/what /mnt/where -o commit_interval=5
> > and
> > mount /mnt/where -o remount,commit_interval=3000
>
> I just implemented that on the way back from OLS, will check it in
> shortly.
Nice... with -o remount possibility it might be even usefull for
noflushd.
Hmm....
Does mount /dev/hda1 -o remount,commit_interval=3000 work? If no we'll
have some ugly parsing of /proc/mounts to be added into noflushd...
Pavel
--
(about SSSCA) "I don't say this lightly. However, I really think that the U.S.
no longer is classifiable as a democracy, but rather as a plutocracy." --hpa