The issue isn't the 4k blocks. The issues are the 512 bytes constants
or 1 << 9 or DEV_BSHIFT or DEV_BSIZE that are all over the tree. I
know for a fact that softraid is busted with these devices and you can
bet that the wrong assumption has been made in other places as well.
Making these drives work will be painful. But we could start with the
520 byte ones ;-)
On Thu, Dec 17, 2009 at 01:36:34PM +1100, Aaron Mason wrote:
> On Thu, Dec 17, 2009 at 1:28 PM, Kenneth Westerback
> <kwesterb...@gmail.com> wrote:
> > On Wed, Dec 16, 2009 at 9:21 PM, Aaron Mason <simplersolut...@gmail.com>
> wrote:
> >> On Thu, Dec 17, 2009 at 12:46 PM, Kenneth Westerback
> >> <kwesterb...@gmail.com> wrote:
> >>> On Wed, Dec 16, 2009 at 7:12 PM, Marco Peereboom <sl...@peereboom.us>
> wrote:
> >>>> yes
> >>>>
> >>>> On Thu, Dec 17, 2009 at 12:30:46AM +0100, Robert wrote:
> >>>>> Hi,
> >>>>>
> >>>>> I just read [1,2] that Western Digital (and probably others) will start
> >>>>> to sell disks with an internal block size of 4KB instead of 512 byte.
> >>>>>
> >>>>> The article mentions that this might lead to a considerable performance
> >>>>> impact if the logical partition alignment is not in sync with the
> >>>>> physical one ("e.g. if partitions starts at sector 63 wich is not a
> >>>>> multiple of 4KB" [1]):
> >>>>> "If a block of 4KB should be written, which is (logically) aligned, but
> >>>>> physically in fact 2 sectors, then both physical sectors will need to
> be
> >>>>> read, partially modified and then written back which leads to a serious
> >>>>> performance hit." [1]
> >>>>>
> >>>>> Any comments from someone who has a good knowledge of this area? Will
> >>>>> this affect OpenBSD?
> >>>>>
> >>>>> regards,
> >>>>> Robert
> >>>>>
> >>>>> [1] (german)
> >>>>>
> >>>
> http://www.heise.de/newsticker/meldung/Festplatten-mit-4-KByte-Sektorgroesse-
> >>> 887759.html
> >>>>> [2]
> http://www.wdc.com/wdproducts/library/whitepapers/en/2579-771430.pdf
> >>>>
> >>>>
> >>>
> >>> 4K sectors themselves should work ok as we already do that for ffs on
> >>> CD media. Which works the last I checked. Ensuring alignment on blocks
> >>> will have to be checked. It sounds like some should donate a few of
> >>> these 4K devices to the project so we can fix any issues asap.
> >>>
> >>> .... Ken
> >>>
> >>>
> >>
> >> I'm looking into the possibility of modifying QEMU for testing this.
> >> Not sure how close such things would be to reality, but it's worth a
> >> look in any case and would drop the cost of development significantly.
> >>
> >> --
> >> Aaron Mason - Programmer, open source addict
> >> I've taken my software vows - for beta or for worse
> >>
> >
> > If you just want to see what you can do with/to 4K sectors, read the
> > vnconfig man page (-t option) and create yourself some 4K sector
> > devices in your disktab.
> >
> > However without hardware testing you never know what the clever dicks
> > at the vendor have done. :-).
> >
> > .... Ken
> >
>
> That's a good point.
>
> From what I've seen, 4k blocks are supported by most filesystems
> anyway - and besides, provided the partitions are created on 4k block
> boundaries, there's no reason for any concern IMHO.
>
> I could be wrong - I'm a programmer, not an engineer.
>
> --
> Aaron Mason - Programmer, open source addict
> I've taken my software vows - for beta or for worse
