Messages in this thread | | | Subject | Re: [Ext2-devel] Re: [PATCH 2/3] ext3 reservation allow turn off for specifed file | From | Mingming Cao <> | Date | 26 Oct 2004 16:01:25 -0700 |
| |
On Tue, 2004-10-26 at 13:18, Andrew Morton wrote: > Mingming Cao <cmm@us.ibm.com> wrote: > > > > > I wonder how important this optimisation really is? I bet no applications > > > are using posix_fadvise(POSIX_FADV_RANDOM) anyway. > > > > > I don't know if there is application using the POSIX_FADV_RANDOM. No? If > > this is the truth, I think we don't need this optimization at present. > > Logically reservation does not benefit seeky random write, but there is > > no benchmark showing performance issue so far. We have already provided > > ways for applications turn off reservation through the existing ioctl > > for specified file and -o noreservation mount option for the whole > > filesystem. > > Well we definitely don't want to be encouraging application developers to be > adding ext3-specific ioctls. So we need to work out if any applications > can get significant benefit from manually disabling reservations and if > so, wire up fadvise() into filesystems and do it that way. > Okey. Also, if there is such a need to disable reservation, maybe we could think about closing the reservation window dynamically based on past I/O pattern, when user application did not give any hint.
> Do you know if disabling reservations helps any workloads? > Not that I aware of. We have done several microbenchmark on sequential and random workload, haven't seen regression with reservations so far. But our testing is certainly not comprehensive.
Mingming
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |