Steve Lord wrote:
>
> So I would not in general recommend write caching on a device, unless
> you know enough about it to be satisfied that its cache makes it out
> to disk on power down. Using the rotational power of the spindle to
> generate power to move the head to a special track and flush the cache
> is not unheard of for instance.
It's my understanding (and ardent hope) that modern IDE disks
do this. It's pretty simple - all they need to do when the power
fails is to seek to the vendor area and squirt the write cache
out in one big write - a few tens of milliseconds max. I'll be
taking this up with some manufacturers, see if I can get a definitive
answer.
For more sophisticated storage systems such as RAID controllers,
I guess we'd have to propagate a SCSI write barrier command down to
the controller itself. I'm not sure that the Linux request
and SCSI layers are up to doing that. Have you looked into it?
|